U.S. patent application number 13/589022 was filed with the patent office on 2013-08-29 for methods and compositions of molecular profiling for disease diagnostics.
This patent application is currently assigned to Veracyte, Inc.. The applicant listed for this patent is Bonnie H. Anderson, Darya I. Chudova, Giulia C. Kennedy, Moraima Pagan, Nusrat Rabbee, Eric T. Wang, Hui Wang, Jonathan I. Wilde. Invention is credited to Bonnie H. Anderson, Darya I. Chudova, Giulia C. Kennedy, Moraima Pagan, Nusrat Rabbee, Eric T. Wang, Hui Wang, Jonathan I. Wilde.
Application Number | 20130225662 13/589022 |
Document ID | / |
Family ID | 42170607 |
Filed Date | 2013-08-29 |
United States Patent
Application |
20130225662 |
Kind Code |
A1 |
Kennedy; Giulia C. ; et
al. |
August 29, 2013 |
METHODS AND COMPOSITIONS OF MOLECULAR PROFILING FOR DISEASE
DIAGNOSTICS
Abstract
The present invention relates to compositions, kits, and methods
for molecular profiling and cancer diagnostics, including but not
limited to gene expression product markers, alternative exon usage
markers, and DNA polymorphisms associated with cancer. In
particular, the present invention provides molecular profiles
associated with thyroid cancer, methods of determining molecular
profiles, and methods of analyzing results to provide a
diagnosis.
Inventors: |
Kennedy; Giulia C.; (San
Francisco, CA) ; Anderson; Bonnie H.; (Half Moon Bay,
CA) ; Chudova; Darya I.; (San Jose, CA) ;
Wang; Eric T.; (Milpitas, CA) ; Wang; Hui;
(San Bruno, CA) ; Pagan; Moraima; (San Francisco,
CA) ; Rabbee; Nusrat; (South San Francisco, CA)
; Wilde; Jonathan I.; (Burlingame, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Kennedy; Giulia C.
Anderson; Bonnie H.
Chudova; Darya I.
Wang; Eric T.
Wang; Hui
Pagan; Moraima
Rabbee; Nusrat
Wilde; Jonathan I. |
San Francisco
Half Moon Bay
San Jose
Milpitas
San Bruno
San Francisco
South San Francisco
Burlingame |
CA
CA
CA
CA
CA
CA
CA
CA |
US
US
US
US
US
US
US
US |
|
|
Assignee: |
Veracyte, Inc.
San Francisco
CA
|
Family ID: |
42170607 |
Appl. No.: |
13/589022 |
Filed: |
August 17, 2012 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12592065 |
Nov 17, 2009 |
|
|
|
13589022 |
|
|
|
|
61199585 |
Nov 17, 2008 |
|
|
|
61270812 |
Jul 13, 2009 |
|
|
|
Current U.S.
Class: |
514/44R ;
435/6.12; 435/6.14; 506/9 |
Current CPC
Class: |
A61B 10/0096 20130101;
C12Q 1/6886 20130101; G16B 25/00 20190201; C12Q 2600/178 20130101;
A61B 10/0283 20130101; C12Q 2600/112 20130101; C12Q 2600/158
20130101; G01N 33/574 20130101; A61B 50/30 20160201; G01N 33/57407
20130101; G06Q 99/00 20130101 |
Class at
Publication: |
514/44.R ;
435/6.12; 506/9; 435/6.14 |
International
Class: |
C12Q 1/68 20060101
C12Q001/68 |
Claims
1-78. (canceled)
79. A method of detecting a thyroid condition in a patient
comprising the steps of a. obtaining a sample of thyroid tissue
from said patient; b. assaying a level of expression of one or more
gene expression products in said sample of thyroid tissue; and c.
classifying said sample of thyroid tissue as benign or normal by
applying a trained algorithm to data from step b, wherein said
method has a negative predictive value (NPV) of at least 95% for
independent samples.
80. (canceled)
81. The method of claim 79, wherein said method has a specificity
greater than 70%.
82. The method of claim 79, wherein an overall classification error
rate of said method is less than 6%.
83. The method of claim 79, wherein said sample of thyroid tissue
is obtained by needle aspiration, fine needle aspiration, core
needle biopsy, vacuum assisted biopsy, large core biopsy,
incisional biopsy, excisional biopsy, punch biopsy, shave biopsy,
or skin biopsy.
84. The method of claim 79, wherein said sample of thyroid tissue
comprises thyroid cells and said sample of thyroid tissue is
obtained by fine needle aspiration.
85. The method of claim 79, wherein said gene expression product is
an RNA expression product.
86. The method of claim 85, wherein a level of said RNA expression
product is measured by microarray, SAGE, blotting, RT-PCR,
quantitative PCR, or sequencing.
87. The method of claim 79, wherein said trained algorithm is
trained with at least three training samples, each of which
exhibits a different malignant pathology.
88. The method of claim 79, wherein said trained algorithm is
trained with a training set comprising a training sample with a
pathology selected from the group consisting of: metastatic
melanoma, metastatic renal carcinoma, metastatic breast carcinoma,
and metastatic B cell lymphoma.
89. The method of claim 79, wherein said trained algorithm is
trained with a training set comprising a training sample with a
tissue type selected from the group consisting of: normal thyroid,
follicular adenoma, parathyroid, follicular carcinoma, lymphocytic
thyroiditis, follicular variant papillary thyroid carcinoma,
papillary thyroid carcinoma, nodular hyperplasia, medullary thyroid
carcinoma, Hurthle cell carcinoma, Hurthle cell adenoma, or
anaplastic thyroid carcinoma.
90. The method of claim 79, wherein said trained algorithm is
trained with a training set comprising a training sample obtained
by fine needle aspiration.
91. The method of claim 79, wherein said trained algorithm is
trained with a training set comprising a training sample obtained
by fine needle aspiration and a training sample obtained by
surgical biopsy.
92. The method of claim 79, wherein said classifying is based on a
level of expression in said sample of thyroid tissue of at least
two of the following genes or corresponding Transcript Cluster ID
Nos: CALCA, CEACAM5, SCG3, SCN9A, SYT4, FABP7, FXYD2, HAVCR1,
LOC100101266, NR1H4, PREPL, SLC3A1, DMRT2, GCM2, KIDINS220, KL,
PTH, SYCP2L, AGR3, C10orf81, MYB, SYCP2, VTCN1, CDH19, MLANA,
S100B, SILV, TYR, TYRL, ALDHIB1, CD44, CPE, DEFB1, EGF, FRMD3,
HSD17B6, IGF2BP2, IQCA1, ITGB3, LOC100129258, NUPR1, ODZ1, PVALB,
PVRL2, RPL3, SLC16A1, ST3GAL5, ZBED2, ABCD2, AIM2, ARSG, AUTS2,
C11orf72, C4orf7, CCL19, CCND1, CD36, CD52, CD96, CFH, CFHR1,
CLDN1, CLDN16, CR2, CXCL13, DAB2, DPP4, DPP6, EAF2, EMR3, FABP4,
FBXO2, FLJ42258, FN1, FPR2, FXYD6, GOS2, GABRB2, GIMAP2, GPR174,
GZMK, HCG11, IL7R, ITGB1, KCNA3, KLRG1, LCP1, LIPH, LOC100131599,
LOC647979, LRP1B, MAPK6, MATN2, MDK, MPPED2, MT1F, MT1G, MT1H,
MT1P2, MYEF2, NRCAM, P2RY13, P1GN, PKHDIL1, PLA2G16, PLCB1, PLEK,
PROS1, PTPRC, PTPRE, PYGL, PYHIN1, RGS13, RIMS2, ROS1, RXRG, SCEL,
SEMA3D, SERGEF, SERPINA1, SERPINA2, SLC24A5, SLC34A2, SLC4A4, SNCA,
STK32A, TIMP1, TIMP2, TMSB10, TNFRSF17, WIPI1, ZFYVEI6, ACSL1,
AHNAK, ASPN, BCL2L1, CC2D2B, DDI2, DNAJC13, DYNLT1, FAH, FREM2,
GMFG, HNRNPA3, IGHG1, LRP12, MAGI3, NDUFC2, OR10D1P, PDE8A, PDHA1,
PFKFB2, PHYH, PPP2R2B, PRKG1, PZP, RNF24, RRAGD, SCN2A, SCUBE3,
SDHALP2, SHC1, SLC43A3, THRSP, TNFRSF1A, TXNDC12, or VWA5A.
93. The method of claim 79, wherein said sample of thyroid tissue
has not previously received a definitive diagnosis.
94. The method of claim 79, wherein said sample of thyroid tissue
is subjected to cytological testing that indicates the sample is
ambiguous or suspicious.
95. The method of claim 79, wherein said sample is a
formalin-fixed-paraffin-embedded sample.
96. The method of claim 88, wherein said pathology of said training
sample is metastatic B cell lymphoma.
97. The method of claim 89, wherein said tissue type of said
training sample is selected from the group consisting of:
follicular adenoma, follicular carcinoma, and papillary thyroid
carcinoma.
98. The method of claim 89, wherein said tissue type of said
training sample is Hurthle cell carcinoma.
99. The method of claim 89, wherein said tissue type of said
training sample is selected from the group consisting of:
follicular adenoma, follicular carcinoma, papillary thyroid
carcinoma, Hurthle cell carcinoma, and follicular variant papillary
thyroid carcinoma.
100. The method of claim 79, further comprising treating the
patient on the basis of step (c).
101-135. (canceled)
136. A method of evaluating a sample of thyroid tissue from a
patient comprising the steps of: (a) obtaining said sample of
thyroid tissue from said patient; (b) assaying an expression level
for two or more gene expression products in said sample of thyroid
tissue to generate test data; (c) applying an algorithm to said
expression level of step (b), wherein said algorithm correlates the
expression level of step (b) with expression data obtained from a
plurality of samples, wherein said plurality of samples comprises a
sample obtained by surgical biopsy of thyroid tissue and a sample
obtained by fine needle aspiration of thyroid tissue; and (d) based
on said correlating, (i) identifying said sample of thyroid tissue
as malignant, (ii) identifying said sample of thyroid tissue as
benign, (iii) identifying said sample of thyroid tissue as
non-cancerous, (iv) identifying said sample of thyroid tissue as
non-malignant, or (v) identifying said sample of thyroid tissue as
normal.
137. The method of claim 136, wherein said algorithm is a trained
algorithm trained by said expression data obtained from said
plurality of samples.
138. The method of claim 137, wherein said plurality of samples
comprises at least 200 training samples.
139. The method of claim 137, wherein said plurality of samples
comprises at least 400 training samples.
140. The method of claim 137, wherein one or more of said plurality
of samples has a pathology selected from the group consisting of:
metastatic melanoma, metastatic renal carcinoma, metastatic breast
carcinoma, and metastatic B cell lymphoma.
141. The method of claim 137, wherein one or more of said plurality
of samples has a pathology selected from the group consisting of:
normal thyroid, follicular adenoma, parathyroid, follicular
carcinoma, lymphocytic thyroiditis, follicular variant papillary
thyroid carcinoma, papillary thyroid carcinoma, nodular
hyperplasia, medullary thyroid carcinoma, Hurthle cell carcinoma,
Hurthle cell adenoma, or anaplastic thyroid carcinoma.
142. The method of claim 137, wherein said trained algorithm is
trained by correlating a gene expression profile of a first
sub-type of thyroid tissue with a gene expression profile of at
least six sub-types of thyroid tissue that are not of said first
sub-type.
143. The method of claim 137, wherein said sample of thyroid tissue
is obtained by needle aspiration, fine needle aspiration, core
needle biopsy, vacuum assisted biopsy, large core biopsy,
incisional biopsy, excisional biopsy, punch biopsy, shave biopsy,
or skin biopsy.
144. The method of claim 137, wherein said sample of thyroid tissue
is subjected to cytological testing, a result of which is
indeterminate.
145. The method of claim 137, wherein said gene expression product
is an RNA gene expression product.
146. The method of claim 137, wherein said correlating uses gene
expression data of at least two of the following genes or
corresponding Transcript Cluster ID Nos: CALCA, CEACAM5, SCG3,
SCN9A, SYT4, FABP7, FXYD2, HAVCR1, LOC100101266, NR1H4, PREPL,
SLC3A1, DMRT2, GCM2, KIDINS220, KL, PTH, SYCP2L, AGR3, C10orf81,
MYB, SYCP2, VTCN1, CDH19, MLANA, S100B, SILV, TYR, TYRL, ALDHIB1,
CD44, CPE, DEFB1, EGF, FRMD3, HSD17B6, IGF2BP2, IQCA1, ITGB3,
LOC100129258, NUPR1, ODZ1, PVALB, PVRL2, RPL3, SLC16A1, ST3GAL5,
ZBED2, ABCD2, AIM2, ARSG, AUTS2, C11orf72, C4orf7, CCL19, CCND1,
CD36, CD52, CD96, CFH, CFHR1, CLDN1, CLDN16, CR2, CXCL13, DAB2,
DPP4, DPP6, EAF2, EMR3, FABP4, FBXO2, FLJ42258, FN1, FPR2, FXYD6,
GOS2, GABRB2, GIMAP2, GPR174, GZMK, HCG11, IL7R, ITGB1, KCNA3,
KLRG1, LCP1, LIPH, LOC100131599, LOC647979, LRP1B, MAPK6, MATN2,
MDK, MPPED2, MT1F, MT1G, MT1H, MT1P2, MYEF2, NRCAM, P2RY13, PIGN,
PKHD1L1, PLA2G16, PLCB1, PLEK, PROS1, PTPRC, PTPRE, PYGL, PYHIN1,
RGS13, RIMS2, ROS1, RXRG, SCEL, SEMA3D, SERGEF, SERPINA1, SERPINA2,
SLC24A5, SLC34A2, SLC4A4, SNCA, STK32A, TIMP1, TIMP2, TMSB10,
TNFRSFI7, WIPI1, ZFYVE16, ACSL1, AHNAK, ASPN, BCL2L1, CC2D2B, DDI2,
DNAJCI3, DYNLT1, FAH, FREM2, GMFG, HNRNPA3, IGHG1, LRP12, MAGI3,
NDUFC2, OR10D1P, PDE8A, PDHA1, PFKFB2, PHYH, PPP2R2B, PRKG1, PZP,
RNF24, RRAGD, SCN2A, SCUBE3, SDHALP2, SHC1, SLC43A3, THRSP,
TNFRSF1A, TXNDC12, or VWA5A.
147. The method of claim 137, wherein said sample of thyroid tissue
is subjected to cytological testing that indicates the sample is
ambiguous or suspicious.
148. The method of claim 137, further comprising treating the
patient on the basis of step (d).
149. A method of detecting a thyroid condition in a patient
comprising the steps of: a. obtaining a sample of thyroid tissue
from said patient; b. assaying a level of expression of one or more
gene expression products in said sample of thyroid tissue; and c.
classifying said sample of thyroid tissue as benign by applying an
algorithm to data from step b, wherein said algorithm correlates
said data from step b with expression data obtained from a
plurality of samples, wherein said plurality of samples comprises a
sample with a pathology that is a metastatic cancer from a
non-thyroid organ.
150. The method of claim 149, wherein said algorithm is a trained
algorithm trained with expression data obtained from a training set
comprising a training sample with a pathology selected from the
group consisting of: metastatic melanoma, metastatic renal
carcinoma, metastatic breast carcinoma, and metastatic B cell
lymphoma.
151. The method of claim 150, wherein said pathology of said
training sample is metastatic B cell lymphoma.
152. The method of claim 149, further comprising treating the
patient on the basis of step (c).
153-159. (canceled)
160. The method of any one of claim 79 or 149, wherein results from
said classifying are sent to a party via a communication
medium.
161. The method of claim 136, wherein results from said identifying
are sent to a party via a communication medium.
162-166. (canceled)
167. The method of claim 79, wherein said method has a negative
predictive value (NPV) of at least 95% for at least 100 independent
samples.
168. The method of claim 79, wherein said method has a sensitivity
of at least 90% for independent samples.
169. The method of claim 79, further comprising treating said
patient on the basis of step c.
170. A method of detecting a thyroid condition in a patient
comprising the steps of: a. obtaining a sample of thyroid tissue
from said patient; b. assaying a level of expression of one or more
gene expression products in said sample of thyroid tissue; and c.
classifying said sample of thyroid tissue as benign or normal by
applying a trained algorithm to data from step b, wherein said
method has a negative predictive value (NPV) of at least 95% for at
least 100 independent samples, and a sensitivity of at least 90 for
at least 100 independent samples.
171. The method of claim 170, wherein said gene expression product
is an RNA expression product.
172. The method of claim 170, wherein said trained algorithm is
trained with at least three training samples, each of which
exhibits a different malignant pathology.
173. The method of claim 170, wherein said trained algorithm is
trained with a training set comprising a training sample with a
pathology selected from the group consisting of: metastatic
melanoma, metastatic renal carcinoma, metastatic breast carcinoma,
and metastatic B cell lymphoma.
174. The method of claim 170, wherein said trained algorithm is
trained with a training set comprising a training sample with a
tissue type selected from the group consisting of: normal thyroid,
follicular adenoma, parathyroid, follicular carcinoma, lymphocytic
thyroiditis, follicular variant papillary thyroid carcinoma,
papillary thyroid carcinoma, nodular hyperplasia, medullary thyroid
carcinoma, Hurthle cell carcinoma, Hurthle cell adenoma, or
anaplastic thyroid carcinoma.
175. The method of claim 170, wherein said trained algorithm is
trained with a training set comprising a training sample obtained
by fine needle aspiration and a training sample obtained by
surgical biopsy.
176. The method of claim 170, wherein said sample of thyroid tissue
has not previously received a definitive diagnosis.
177. The method of claim 170, wherein said sample of thyroid tissue
is subjected to cytological testing that indicates the sample is
ambiguous or suspicious.
178. The method of claim 173, wherein said pathology of said
training sample is metastatic B cell lymphoma.
179. The method of claim 174, wherein said tissue type of said
training sample is selected from the group consisting of:
follicular adenoma, follicular carcinoma, and papillary thyroid
carcinoma.
180. The method of claim 174, wherein said tissue type of said
training sample is selected from the group consisting of:
follicular adenoma, follicular carcinoma, papillary thyroid
carcinoma, Hurthle cell carcinoma, and follicular variant papillary
thyroid carcinoma.
181. The method of claim 170, further comprising treating said
patient on the basis of step c.
182. The method of claim 170, wherein said sample of thyroid tissue
is obtained by needle aspiration, fine needle aspiration, core
needle biopsy, vacuum assisted biopsy, large core biopsy,
incisional biopsy, excisional biopsy, punch biopsy, shave biopsy,
or skin biopsy
183. The method of claim 170, wherein said sample of thyroid tissue
comprises thyroid cells and said sample of thyroid tissue is
obtained by fine needle aspiration.
184. The method of claim 79, wherein the sample is classified as
not benign or not normal, and further comprising treating the
patient.
185. The method of claim 79 or 170, wherein prior to step (b), a
cytological analysis of the tissue is performed, wherein if the
tissue is ambiguous based on the cytological analysis, then steps
(b) and (c) are performed.
186. The method of claim 79 or 170, wherein prior to step (b), a
cytological analysis of the tissue is performed, wherein (i) if the
tissue is ambiguous based on the cytological analysis, then steps
(b) and (c) are performed; and (ii) if the tissue is malignant or
benign then steps (b) and (c) are not performed.
Description
CROSS REFERENCE
[0001] This application claims the benefit of U.S. Provisional
Application No. 61/199,585, entitled "Methods and Compositions of
Molecular Profiling for Diagnosis of Cancer" filed Nov. 17, 2008,
and U.S. Provisional Application No. 61/270,812, entitled "Methods
and Compositions of Molecular Profiling for Diagnosis of Cancer"
filed Jul. 13, 2009, both of which are incorporated herein by
reference in its entirety.
BACKGROUND OF THE INVENTION
[0002] Cancer is the second leading cause of death in the United
States and one of the leading causes of mortality worldwide. Nearly
25 million people are currently living with cancer, with 11 million
new cases diagnosed each year. Furthermore, as the general
population continues to age, cancer will become a bigger and bigger
problem. The World Health Organization projects that by the year
2020, global cancer rates will increase by 50%.
[0003] Successful treatment of cancer starts with early and
accurate diagnosis. Current methods of diagnosis include
cytological examination of tissue samples taken by biopsy or
imaging of tissues and organs for evidence of aberrant cellular
proliferation. While these techniques have proven to be both useful
and inexpensive, they suffer from a number of drawbacks. First,
cytological analysis and imaging techniques for cancer diagnosis
often require a subjective assessment to determine the likelihood
of malignancy. Second, the increased use of these techniques has
lead to a sharp increase in the number of indeterminate results in
which no definitive diagnosis can be made. Third, these routine
diagnostic methods lack a rigorous method for determining the
probability of an accurate diagnosis. Fourth, these techniques may
be incapable of detecting a malignant growth at very early stages.
Fifth, these techniques do not provide information regarding the
basis of the aberrant cellular proliferation.
[0004] Many of the newer generation of treatments for cancer, while
exhibiting greatly reduced side effects, are specifically targeted
to a certain metabolic or signaling pathway, and will only be
effective against cancers that are reliant on that pathway.
Further, the cost of any treatments can be prohibitive for an
individual, insurance provider, or government entity. This cost
could be at least partially offset by improved methods that
accurately diagnose cancers and the pathways they rely on at early
stages. These improved methods would be useful both for preventing
unnecessary therapeutic interventions as well as directing
treatment.
[0005] In the case of thyroid cancer it is estimated that out of
the approximately 130,000 thyroid removal surgeries performed each
year due to suspected malignancy in the United States, only about
54,000 are necessary. Thus, approximately 76,000 unnecessary
surgeries are performed annually. In addition, there are continued
treatment costs and complications due to the need for lifelong drug
therapy to replace the lost thyroid function. Accordingly, there is
a need for improved testing modalities and business practices that
improve upon current methods of cancer diagnosis.
[0006] The thyroid has at least two kinds of cells that make
hormones. Follicular cells make thyroid hormone, which affects
heart rate, body temperature, and energy level. C cells make
cacitonin, a hormone that helps control the level of calcium in the
blood. Abnormal growth in the thyroid can result in the formation
of nodules, which can be either benign or malignant. Thyroid cancer
includes at least four different kinds of malignant tumors of the
thyroid gland: papillary, follicular, medullary and anaplastic.
SUMMARY OF THE INVENTION
[0007] The present invention includes a method for diagnosing
thyroid disease in a subject, the method comprising (a) providing a
nucleic acid sample from a subject; (b) detecting the amount of one
or more genes, gene products, or transcripts selected from the
group consisting of the genes or transcripts listed in Tables 2 or
their complement; and (c) determining whether said subject has or
is likely to have a malignant or benign thyroid condition based on
the results of step (b).
[0008] The present invention also includes a composition comprising
one or more binding agents that specifically bind to the one or
more polymorphisms selected from the group consisting of the
polymorphisms listed in Tables.
INCORPORATION BY REFERENCE
[0009] All publications and patent applications mentioned in this
specification are herein incorporated by reference to the same
extent as if each individual publication or patent application was
specifically and individually indicated to be incorporated by
reference.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The novel features of the invention are set forth with
particularity in the appended claims. A better understanding of the
features and advantages of the present invention will be obtained
by reference to the following detailed description that sets forth
illustrative embodiments, in which the principles of the invention
are utilized, and the accompanying drawings of which:
[0011] FIG. 1 is a table listing 75 thyroid samples examined for
gene expression analysis using the Affymetrix Human Exon 10ST array
to identify genes that are significantly differentially expressed
or alternatively spliced between malignant, benign, and normal
samples. The name for each sample and the pathological
classification is listed.
[0012] FIG. 2 is a table listing the top 100 differentially
expressed genes at the gene level. Data are from the dataset in
which benign malignant and normal thyroid samples were compared at
the gene level. Markers were selected based on statistical
significance after Benjamini and Hochberg correction for false
discover rate (FDR). Positive numbers denote up regulation and
negative numbers denote down regulation of expression.
[0013] FIG. 3 is a table listing the top 100 alternatively spliced
genes. Data are from the dataset in which benign malignant and
normal thyroid samples were compared at the gene level. Markers
were selected based on statistical significance after Benjamini and
Hochberg correction for false discovery rate (FDR).
[0014] FIG. 4 is a table listing the top 100 differentially
expressed genes at the probe-set level. Data were from the
Probe-set dataset. Positive numbers denote up-regulation of gene
expression, while negative numbers denote down regulation.
[0015] FIG. 5 is a table listing the top 100 significant diagnostic
markers determined by gene level analysis. Markers in this list
show both differential gene expression and alternative exon
splicing. Positive numbers denote up-regulation, while negative
numbers denote down regulation. This table lists 3-sets of
calculated fold-changes for any given marker to allow comparison
between the groups malignant vs. benign, benign, versus normal, and
malignant versus normal.
[0016] FIG. 6 is a table listing the genes identified as
contributing to thyroid cancer diagnosis by molecular profiling of
gene expression levels and/or alternative exon splicing. Markers
identified from the dataset in which benign, malignant and normal
samples were analyzed at the gene level are referred to as BMN in
the data source column; and likewise, markers identified from
dataset in which the benign and malignant samples were analyzed at
the gene level are referred to as BM in the data source column.
Similarly, markers identified at the probe-set level from the
dataset in which benign, and malignant samples were analyzed are
referred to as Probe-set in the data source column.
[0017] FIG. 7 is a table listing tissue samples examined for gene
expression analysis. The samples were classified by pathological
analysis as benign (B) or malignant (M). Benign samples were
further classified as follicular adenoma (FA), lymphocytic
thyroiditis (LCT), or nodular hyperplasia (NHP). Malignant samples
were further classified as Hurthle cell carcinoma (HC), follicular
carcinoma (FC), follicular variant of pappillary thyroid carcinoma
(FVPTC), papillary thyroid carcinoma (PTC), medullary thyroid
carcinoma (MTC), or anaplastic carcinoma (ATC).
[0018] FIG. 8 is a table listing fine needle aspirate samples
examined for gene expression analysis. The samples were classified
by pathological analysis as benign (B) or malignant (M). Benign
samples were further classified as follicular adenoma (FA),
lymphocytic thyroiditis (LCT), Hurthle cell adenoma (HA), or
nodular hyperplasia (NHP). Malignant samples were further
classified as Hurthle cell carcinoma (HC), follicular carcinoma
(FC), follicular variant of pappillary thyroid carcinoma (FVPTC),
papillary thyroid carcinoma (PTC), medullary thyroid carcinoma
(MTC), or anaplastic carcinoma (ATC).
[0019] FIG. 9 is a table listing genes identified from expression
analysis of the tissue samples listed in FIG. 7 which exhibit
significant differences in expression between malignant and benign
samples as determined by feature selection using LIMMA (linear
models for micro array data) and SVM (support vector machine) for
classification of malignant vs. benign samples. Rank denotes the
marker significance (lower rank, higher significance) after
Benjamini and Hochberg correction for False Discovery Rate (FDR).
Gene symbol denotes the name of the gene. TCID denotes the
transcript cluster ID of the gene used in the Affymetrix Human Exon
10ST array. Ref Seq denotes the name of the corresponding reference
sequence for that gene. The column labeled "Newly Discovered
Marker" denotes gene expression markers which have not previously
been described as differentially expressed in malignant vs. benign
thyroid tissues.
[0020] FIG. 10 is a table listing genes identified from expression
analysis of the tissue samples listed in FIG. 8 which exhibit
significant differences in expression between medullary thyroid
carcinoma (MTC) and other pathologies as determined by feature
selection using LIMMA (linear models for micro array data) and SVM
(support vector machine) for classification of MTC vs. other
samples. Rank denotes the marker significance (lower rank, higher
significance) after Benjamini and Hochberg correction for False
Discovery Rate (FDR). Gene symbol denotes the name of the gene.
TCID denotes the transcript cluster ID of the gene used in the
Affymetrix Human Exon 10ST array. P value indicates the statistical
significance of the differential expression between MTC and non-MTC
samples. Fold Change indicates the degree of differential
expression between MTC and non-MTC samples. The column labeled
"Newly Discovered Marker" denotes gene expression markers which
have not previously been described as differentially expressed in
malignant vs. benign thyroid tissues.
[0021] FIG. 11 is a table listing genes identified from expression
analysis of the samples listed in FIGS. 7 and 8 which exhibit
significant differences in expression between benign and malignant
samples as determined by a repeatability based meta-analysis
classification algorithm.
[0022] FIG. 12 is a table listing genes identified from expression
analysis of the samples listed in FIGS. 7 and 8 which exhibit
significant (posterior probability>0.9) differences in
expression between benign and malignant samples as determined by
Bayesian ranking of the differentially expressed genes. deriving
type I and type II error rates from previously published studies to
determine prior probabilities, combining these prior probabilities
with the output of the dataset derived from expression analysis of
the samples listed in FIG. 10 to estimate posterior probabilities
of differential gene expression, and then combining the results of
the expression analysis of the samples listed in FIG. 11 with the
estimated posterior probabilities to calculate final posterior
probabilities of differential gene expression. These posterior
probabilities were then used to rank the differentially expressed
genes.
[0023] FIG. 13 is a table listing genes identified from expression
analysis of the samples listed in FIG. 7 which exhibit differential
expression between samples categorized as FA, LCT, NHP, HC, FC,
FVPTC, PTC, MTC, or ATC as determined by feature selection using
LIMMA (linear models for micro array data) and SVM (support vector
machine) for classification.
[0024] FIG. 14 is a table listing fine needle aspirate samples
examined for micro RNA (miRNA) expression analysis using an Agilent
Human v2 miRNA microarray chip. The samples were classified by
pathological analysis as benign (B) or malignant (M). Benign
samples were further classified as follicular adenoma (FA), or
nodular hyperplasia (NHP). Malignant samples were further
classified as follicular carcinoma (FC), follicular variant of
pappillary thyroid carcinoma (FVPTC), papillary thyroid carcinoma
(PTC), or medullary thyroid carcinoma (MTC).
[0025] FIG. 15 is a table listing fine needle aspirate samples
examined for micro RNA (miRNA) expression analysis using an
Illumina Human v2 miRNA array. The samples were classified by
pathological analysis as benign (B), non diagnostic, or malignant
(M). Benign samples were further classified as benign nodule (BN),
follicular neoplasm (FN), (LCT), or (NHP). Malignant samples were
further classified as (FVPTC), or (PTC).
[0026] FIG. 16 is a table listing micro RNAs (miRNAs) identified
from analysis of the samples listed in FIG. 14 which exhibit
differential expression between samples categorized as benign or
malignant. The miRNA column denotes the name of the miRNA. The CHR
column denotes the chromosome the miRNA is located on. The P column
denotes the statistical confidence or p-value provided by the
analysis. The DE column denotes whether the listed miRNA is
upregulated (1) in malignant samples or downregulated (-1) in
malignant samples. The patent column denotes any patents or
applications that describe these miRNAs.
[0027] FIG. 17 is a table listing micro RNAs (miRNAs) identified
from analysis of the samples listed in FIG. 15 which exhibit
differential expression between samples categorized as benign or
malignant. The miRNA column denotes the name of the miRNA. The
probe ID column denotes the corresponding probe ID in the illumina
array. The CHR column denotes the chromosome the miRNA is located
on. The P column denotes the statistical confidence or p-value
provided by the analysis. The DE column denotes whether the listed
miRNA is upregulated (no sign) in malignant samples or
downregulated (negative sign) in malignant samples. The Rep column
denotes the repeatability score provided by a "hot probes" type
analysis of the hybridization data. The patent column denotes any
patents or applications that describe these miRNAs.
[0028] FIG. 18 is a flow chart describing how molecular profiling
may be used to improve the accuracy of routine cytological
examination. FIG. 18A and FIG. 18B describe alternate embodiments
of the molecular profiling business.
[0029] FIG. 19 is an illustration of a kit provided by the
molecular profiling business.
[0030] FIG. 20 is an illustration of a molecular profiling results
report.
[0031] FIG. 21 depicts a computer useful for displaying, storing,
retrieving, or calculating diagnostic results from the molecular
profiling; displaying, storing, retrieving, or calculating raw data
from genomic or nucleic acid expression analysis; or displaying,
storing, retrieving, or calculating any sample or customer
information useful in the methods of the present invention.
[0032] FIG. 22 depicts a titration curve of error rate vs. number
of genes using an SVM-based classification algorithm. The titration
curve plateaus when the classification algorithm examines 200-250
genes. These data indicate that the overall error rate of the
current algorithm was 4% (5/138).
DETAILED DESCRIPTION OF THE INVENTION
I. Introduction
[0033] The present disclosure provides novel methods for diagnosing
abnormal cellular proliferation from a biological test sample, and
related kits and compositions. The present invention also provides
methods and compositions for differential diagnosis of types of
aberrant cellular proliferation such as carcinomas including
follicular carcinomas (FC), follicular variant of papillary thyroid
carcinomas (FVPTC), Hurthle cell carcinomas (HC), Hurthle cell
adenomas (HA); papillary thyroid carcinomas (PTC), medullary
thyroid carcinomas (MTC), and anaplastic carcinomas (ATC); adenomas
including follicular adenomas (FA); nodule hyperplasias (NHP);
colloid nodules (CN); benign nodules (BN); follicular neoplasms
(FN); lymphocytic thyroiditis (LCT), including lymphocytic
autoimmune thyroiditis; parathyroid tissue; renal carcinoma
metastasis to the thyroid; melanoma metastasis to the thyroid;
B-cell lymphoma metastasis to the thyroid; breast carcinoma to the
thyroid; benign (B) tumors, malignant (M) tumors, and normal (N)
tissues. The present invention further provides novel markers
including microRNAs (miRNAs) and gene expression product markers
and novel groups of genes and markers useful for the diagnosis,
characterization, and treatment of cellular proliferation.
Additionally the present invention provides business methods for
providing enhanced diagnosis, differential diagnosis, monitoring,
and treatment of cellular proliferation.
[0034] Cancer is a leading cause of death in the United States.
Early and accurate diagnosis of cancer is critical for effective
management of this disease. It is therefore important to develop
testing modalities and business practices to enable cancer
diagnosis that is more accurately and earlier. Expression product
profiling, also referred to as molecular profiling, provides a
powerful method for early and accurate diagnosis of tumors or other
types of cancers from a biological sample.
[0035] Typically, screening for the presence of a tumor or other
type of cancer, involves analyzing a biological sample taken by
various methods such as, for example, a biopsy. The biological
sample is then prepared and examined by one skilled in the art. The
methods of preparation can include but are not limited to various
cytological stains, and immuno-histochemical methods.
Unfortunately, traditional methods of cancer diagnosis suffer from
a number of deficiencies. These deficiencies include: 1) the
diagnosis may require a subjective assessment and thus be prone to
inaccuracy and lack of reproducibility, 2) the methods may fail to
determine the underlying genetic, metabolic or signaling pathways
responsible for the resulting pathogenesis, 3) the methods may not
provide a quantitative assessment of the test results, and 4) the
methods may be unable to provide an unambiguous diagnosis for
certain samples.
[0036] One hallmark of cancer is dysregulation of normal
transcriptional control leading to aberrant expression of genes or
other RNA transcripts such as miRNAs. Among the aberrantly
expressed transcripts are genes involved in cellular
transformation, for example tumor suppressors and oncogenes. Tumor
suppressor genes and oncogenes may be up-regulated or
down-regulated in tumors when compared to normal tissues. Known
tumor suppressors and oncogenes include, but are not limited to
brca1, brca2, bcr-abl, bcl-2, HER2, N-myc, C-myc, BRAF, RET, Ras,
KIT, Jun, Fos, and p53. This abnormal expression may occur through
a variety of different mechanisms. It is not necessary in the
present invention to understand the mechanism of aberrant
expression, or the mechanism by which carcinogenesis occurs.
Nevertheless, finding a marker or set of markers whose expression
is up or down regulated in a sample as compared to a normal sample
may be indicative of cancer. Furthermore, the particular aberrantly
expressed markers or set of markers may be indicative of a
particular type of cancer, or even a recommended treatment
protocol. Additionally the methods of the present invention are not
meant to be limited solely to canonically defined tumor suppressors
or oncogenes. Rather, it is understood that any marker, gene or set
of genes or markers that is determined to have a statistically
significant correlation with respect to expression level or
alternative gene splicing to a benign, malignant, or normal
diagnosis is encompassed by the present invention.
[0037] In one embodiment, the methods of the present invention seek
to improve upon the accuracy of current methods of cancer
diagnosis. Improved accuracy can result from the measurement of
multiple genes and/or expression markers, the identification of
gene expression products such as miRNAs, rRNA, tRNA and mRNA gene
expression products with high diagnostic power or statistical
significance, or the identification of groups of genes and/or
expression products with high diagnostic power or statistical
significance, or any combination thereof.
[0038] For example, increased expression of a number of receptor
tyrosine kinases has been implicated in carcinogenesis. Measurement
of the gene expression product level of a particular receptor
tyrosine kinase known to be differentially expressed in cancer
cells may provide incorrect diagnostic results leading to a low
accuracy rate. Measurement of a plurality of receptor tyrosine
kinases may increase the accuracy level/by requiring a combination
of alternate expressed genes to occur. In some cases, measurement
of a plurality of genes might therefore increase the accuracy of a
diagnosis by reducing the likelihood that a sample may exhibit an
aberrant gene expression profile by random chance.
[0039] Similarly, some gene expression products within a group such
as receptor tyrosine kinases may be indicative of a disease or
condition when their expression levels are higher or lower than
normal. The measurement of expression levels of other gene products
within that same group may provide diagnostic utility. Thus, in one
embodiment, the invention measures two or more gene expression
products that are within a group. For example, in some embodiments,
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45 or 50
gene expression products are measured from a group. Various groups
are defined within the specification, such as groups useful for
diagnosis of subtypes of thyroid cancer or groups of gene
expression products that fall within particular ontology groups. In
another embodiment, it would be advantageous to measure the
expression levels, of sets of genes that accurately indicate the
presence or absence of cancer from multiple groups. For example,
the invention contemplates the use of 1, 2, 3, 4, 5, 6, 7, 8, 9,
10, 15, 20, 25, 30, 35, 40, 45 or 50 gene expression groups, each
with 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45 or
50 gene expression products measured.
[0040] Additionally, increased expression of other oncogenes such
as for example Ras in a biological sample may also be indicative of
the presence of cancerous cells. In some cases, it may be
advantageous to determine the expression level of several different
classes of oncogenes such as for example receptor tyrosine kinases,
cytoplasmic tyrosine kinases, GTPases, serine/threonine kinases,
lipid kinases, mitogens, growth factors, and transcription factors.
The determination of expression levels and/or exon usage of
different classes or groups of genes involved in cancer progression
may in some cases increase the diagnostic power of the present
invention.
[0041] Groups of expression markers may include markers within a
metabolic or signaling pathway, or genetically or functionally
homologous markers. For example, one group of markers may include
genes involved in the epithelial growth factor signaling pathway.
Another group of markers may include mitogen-activated protein
kinases. The present invention also provides methods and
compositions for detecting (i.e. measuring) measuring gene
expression markers from multiple and/or independent metabolic or
signaling pathways.
[0042] In one embodiment, expression product markers of the present
invention may provide increased accuracy of cancer diagnosis
through the use of multiple expression product markers and
statistical analysis. In particular, the present invention
provides, but is not limited to, RNA expression profiles associated
with thyroid cancers. The present invention also provides methods
of characterizing thyroid tissue samples, and kits and compositions
useful for the application of said methods. The disclosure further
includes methods for running a molecular profiling business.
[0043] The present disclosure provides methods and compositions for
improving upon the current state of the art for diagnosing
cancer.
[0044] In some embodiments, the present invention provides a method
of diagnosing cancer comprising the steps of: obtaining a
biological sample comprising gene expression products; determining
the expression level for one or more gene expression products of
the biological sample; and identifying the biological sample as
cancerous wherein the gene expression level is indicative of the
presence of thyroid cancer in the biological sample. This can be
done by correlating the gene expression levels with the presence of
thyroid cancer in the biological sample. In one embodiment, the
gene expression products are selected from FIG. 6. In some
embodiments, the method further comprises the step of comparing the
expression level of the one or more gene expression products to a
control expression level for each gene expression product in a
control sample, wherein the biological sample is identified as
cancerous if there is a difference in the gene expression level
between a gene expression product in the biological sample and the
control sample.
[0045] In some embodiments, the present invention provides a method
of diagnosing cancer comprising the steps of: obtaining a
biological sample comprising alternatively spliced gene expression
products; determining the expression level for one or more gene
expression products of the biological sample; and identifying the
biological sample as cancerous wherein the gene expression level is
indicative of the presence of thyroid cancer in the biological
sample. This can be done by correlating the gene expression levels
with the presence of thyroid cancer in the biological sample. In
one embodiment, the alternatively spliced gene expression products
are selected from FIG. 6, wherein the differential gene expression
product alternative exon usage is compared between the biological
sample and a control sample; and identifying the biological sample
as cancerous if there is a difference in gene expression product
alternative exon usage between the biological sample and the
control sample at a specified confidence level. In some
embodiments, the genes selected from FIG. 6 are further selected
from genes listed in FIG. 2, FIG. 3, FIG. 4, or FIG. 5.
[0046] In some embodiments, the present invention provides a method
of diagnosing cancer that gives a specificity or sensitivity that
is greater than 70% using the subject methods described herein,
wherein the gene expression product levels are compared between the
biological sample and a control sample; and identifying the
biological sample as cancerous if there is a difference in the gene
expression levels between the biological sample and the control
sample at a specified confidence level. In some embodiments, the
specificity and/or sensitivity of the present method is at least
70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%,
95%, 96%, 97%, 98%, 99%, or more.
[0047] In some embodiments, the nominal specificity is greater than
or equal to 70%. The nominal negative predictive value (NPV) is
greater than or equal to 95%. In some embodiments, the NPV is at
least 95%, 95.5%, 96%, 96.5%, 97%, 97.5%, 98%, 98.5%, 99%, 99.5% or
more.
[0048] Sensitivity typically refers to TP/(TP+FN), where TP is true
positive and FN is false negative. Number of Continued
Indeterminate results divided by the total number of malignant
results based on adjudicated histopathology diagnosis. Specificity
typically refers to TN/(TN+FP), where TN is true negative and FP is
false positive. The number of benign results divided by the total
number of benign results based on adjudicated histopathology
diagnosis. Positive Predictive Value (PPV): TP/(TP+FP); Negative
Predictive Value (NPV): TN/(TN+FN).
[0049] Marker panels are chosen to accommodate adequate separation
of benign from non-benign expression profiles. Training of this
multi-dimensional classifier, i.e., algorithm, was performed on
over 500 thyroid samples, including >300 thyroid FNAs. Many
training/test sets were used to develop the preliminary algorithm.
An exemplary data set is shown in FIG. 22. First the overall
algorithm error rate is shown as a function of gene number for
benign vs non-benign samples. All results are obtained using a
support vector machine model which is trained and tested in a
cross-validated mode (30-fold) on the samples.
[0050] In some embodiments, the difference in gene expression level
is at least 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45% or 50% or more.
In some embodiments, the difference in gene expression level is at
least 2, 3, 4, 5, 6, 7, 8, 9, 10 fold or more. In some embodiments,
the biological sample is identified as cancerous with an accuracy
of greater than 75%, 80%, 85%, 90%, 95%, 99% or more. In some
embodiments, the biological sample is identified as cancerous with
a sensitivity of greater than 95%. In some embodiments, the
biological sample is identified as cancerous with a specificity of
greater than 95%. In some embodiments, the biological sample is
identified as cancerous with a sensitivity of greater than 95% and
a specificity of greater than 95%. In some embodiments, the
accuracy is calculated using a trained algorithm.
[0051] In some embodiments, the present invention provides gene
expression products corresponding to genes selected from Table 3,
Table 4 and/or Table 5.
[0052] In some embodiments, the present invention provides a method
of diagnosing cancer comprising using gene expression products from
one or more of the following signaling pathways. The signaling
pathways from which the genes can be selected include but are not
limited to: acute myeloid leukemia signaling, somatostatin receptor
2 signaling, cAMP-mediated signaling, cell cycle and DNA damage
checkpoint signaling, G-protein coupled receptor signaling,
integrin signaling, melanoma cell signaling, relaxin signaling, and
thyroid cancer signaling. In some embodiments, more than one gene
is selected from a single signaling pathway to determine and
compare the differential gene expression product level between the
biological sample and a control sample. Other signaling pathways
include, but are not limited to, an adherens, ECM, thyroid cancer,
focal adhesion, apoptosis, p53, tight junction, TGFbeta, ErbB, Wnt,
pathways in cancer overview, cell cycle, VEGF, Jak/STAT, MAPK,
PPAR, mTOR or autoimmune thyroid pathway. In other embodiments, at
least two genes are selected from at least two different signaling
pathways to determine and compare the differential gene expression
product level between the biological sample and the control sample.
Methods and compositions of the invention can have genes selected
from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50
or more signaling pathways and can have from 1, 2, 3, 4, 5, 6, 7,
8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more gene expression
products from each signaling pathway, in any combination. In some
embodiments, the set of genes combined give a specificity or
sensitivity of greater than 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%,
90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 99.5%, or a
positive predictive value or negative predictive value of at least
95%, 95.5%, 96%, 96.5%, 97%, 97.5%, 98%, 98.5%, 99%, 99.5% or
more.
[0053] In some embodiments, the present invention provides a method
of diagnosing cancer comprising genes selected from at least two
different ontology groups. In some embodiments, the ontology groups
from which the genes can be selected include but are not limited
to: cell aging, cell cortex, cell cycle, cell death/apoptosis, cell
differentiation, cell division, cell junction, cell migration, cell
morphogenesis, cell motion, cell projection, cell proliferation,
cell recognition, cell soma, cell surface, cell surface linked
receptor signal transduction, cell adhesion, transcription, immune
response, or inflammation. In some embodiments, more than one gene
is selected from a single ontology group to determine and compare
the differential gene expression product level between the
biological sample and a control sample. In other embodiments, at
least two genes are selected from at least two different ontology
groups to determine and compare the differential gene expression
product level between the biological sample and the control sample.
Methods and compositions of the invention can have genes selected
from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50
or more gene ontology groups and can have from 1, 2, 3, 4, 5, 6, 7,
8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more gene expression
products from each gene ontology group, in any combination. In some
embodiments, the set of genes combined give a specificity or
sensitivity of greater than 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%,
90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 99.5%, or a
positive predictive value or negative predictive value of at least
95%, 95.5%, 96%, 96.5%, 97%, 97.5%, 98%, 98.5%, 99%, 99.5% or
more.
[0054] In some embodiments, the present invention provides a method
of classifying cancer comprising the steps of: obtaining a
biological sample comprising gene expression products; determining
the expression level for one or more gene expression products of
the biological sample that are differentially expressed in
different subtypes of a cancer; and identifying the biological
sample as cancerous wherein the gene expression level is indicative
for a subtype of cancer. In some embodiments, the method further
comprises the step of comparing the expression level of the one or
more gene expression products to a control expression level for
each gene expression product in a control sample, wherein the
biological sample is identified as cancerous if there is a
difference in the gene expression level between a gene expression
product in the biological sample and the control sample. In some
embodiments, the subject methods distinguish follicular carcinoma
from medullary carcinoma. In some embodiments, the subject methods
distinguish a benign thyroid disease from a malignant thyroid
tumor/carcinoma.
[0055] In some embodiments, the gene expression product of the
subject methods is a protein, and the amount of protein is
compared. The amount of protein can be determined by one or more of
the following: ELISA, mass spectrometry, blotting, or
immunohistochemistry. RNA can be measured by one or more of the
following: microarray, SAGE, blotting, RT-PCR, or quantitative
PCR.
[0056] In some embodiments, the difference in gene expression
level, for example, mRNA, protein, or alternatively spliced gene
product, between a biological sample and a control sample that can
be used to diagnose cancer is at least 1.5, 2, 2.5, 3, 3.5, 4, 4.5,
5, 5.5, 6, 6.5, 7, 7.5, 8, 8.5, 9, 9.5, 10 fold or more.
[0057] In some embodiments, the biological sample is classified as
cancerous or positive for a subtype of cancer with an accuracy of
greater than 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%,
94%, 95%, 96%, 97%, 98%, 99%, or 99.5%. The diagnosis accuracy as
used herein includes specificity, sensitivity, positive predictive
value, negative predictive value, and/or false discovery rate.
[0058] When classifying a biological sample for diagnosis of
cancer, there are typically four possible outcomes from a binary
classifier. If the outcome from a prediction is p and the actual
value is also p, then it is called a true positive (TP); however if
the actual value is n then it is said to be a false positive (FP).
Conversely, a true negative has occurred when both the prediction
outcome and the actual value are n, and false negative is when the
prediction outcome is n while the actual value is p. In one
embodiment, consider a diagnostic test that seeks to determine
whether a person has a certain disease. A false positive in this
case occurs when the person tests positive, but actually does not
have the disease. A false negative, on the other hand, occurs when
the person tests negative, suggesting they are healthy, when they
actually do have the disease. In some embodiments, ROC curve
assuming real-world prevalence of subtypes can be generated by
re-sampling errors achieved on available samples in relevant
proportions.
[0059] The positive predictive value (PPV), or precision rate, or
post-test probability of disease, is the proportion of patients
with positive test results who are correctly diagnosed. It is the
most important measure of a diagnostic method as it reflects the
probability that a positive test reflects the underlying condition
being tested for. Its value does however depend on the prevalence
of the disease, which may vary. In one'example, FP (false
positive); TN (true negative); TP (true positive); FN (false
negative).
False positive rate(.alpha.)=FP/(FP+TN)-specificity
False negative rate(.beta.)=FN/(TP+FN)-sensitivity
Power=sensitivity=1-.beta.
Likelihood-ratio positive=sensitivity/(1-specificity)
Likelihood-ratio negative=(1-sensitivity)/specificity
[0060] The negative predictive value is the proportion of patients
with negative test results who are correctly diagnosed. PPV and NPV
measurements can be derived using appropriate disease subtype
prevalence estimates. An estimate of the pooled malignant disease
prevalence can be calculated from the pool of indeterminates which
roughly classify into B vs M by surgery. For subtype specific
estimates, in some embodiments, disease prevalence may sometimes be
incalculable because there are not any available samples. In these
cases, the subtype disease prevalence can be substituted by the
pooled disease prevalence estimate.
[0061] In some embodiments, the level of expression products or
alternative exon usage is indicative of one of the following:
follicular cell carcinoma, anaplastic carcinoma, medullary
carcinoma, or sarcoma. In some embodiments, the one or more genes
selected using the methods of the present invention for diagnosing
cancer contain representative sequences corresponding to a set of
metabolic or signaling pathways indicative of cancer.
[0062] In some embodiments, the results of the expression analysis
of the subject methods provide a statistical confidence level that
a given diagnosis is correct. In some embodiments, such statistical
confidence level is above 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%,
97%, 98%, 99% or 99.5%.
[0063] In another aspect, the present invention provides a
composition for diagnosing cancer comprising oligonucleotides
comprising a portion of one or more of the genes listed in FIG. 6
or their complement, and a substrate upon which the
oligonucleotides are covalently attached. The composition of the
present invention is suitable for use in diagnosing cancer at a
specified confidence level using a trained algorithm. In one
example, the composition of the present invention is used to
diagnose thyroid cancer.
[0064] In one aspect of the present disclosure, samples that have
been processed by a cytological company, subjected to routine
methods and stains, diagnosed and categorized, are then subjected
to molecular profiling as a second diagnostic screen. This second
diagnostic screen enables: 1) a significant reduction of false
positives and false negatives, 2) a determination of the underlying
genetic, metabolic, or signaling pathways responsible for the
resulting pathology, 3) the ability to assign a statistical
probability to the accuracy of the diagnosis, 4) the ability to
resolve ambiguous results, and 5) the ability to distinguish
between sub-types of cancer.
[0065] For example, in the specific case of thyroid cancer,
molecular profiling of the present invention may further provide a
diagnosis for the specific type of thyroid cancer (e.g. papillary,
follicular, medullary, or anaplastic). The results of the molecular
profiling may further allow one skilled in the art, such as a
scientist or medical professional to suggest or prescribe a
specific therapeutic intervention. Molecular profiling of
biological samples may also be used to monitor the efficacy of a
particular treatment after the initial diagnosis. It is further
understood that in some cases, molecular profiling may be used in
place of, rather than in addition to, established methods of cancer
diagnosis.
[0066] In one aspect, the present invention provides algorithms and
methods that can be used for diagnosis and monitoring of a genetic
disorder. A genetic disorder is an illness caused by abnormalities
in genes or chromosomes. While some diseases, such as cancer, are
due in part to genetic disorders, they can also be caused by
environmental factors. In some embodiments, the algorithms and the
methods disclosed herein are used for diagnosis and monitoring of a
cancer such as thyroid cancer.
[0067] Genetic disorders can be typically grouped into two
categories: single gene disorders and multifactorial and polygenic
(complex) disorders. A single gene disorder is the result of a
single mutated gene. There are estimated to be over 4000 human
diseases caused by single gene defects. Single gene disorders can
be passed on to subsequent generations in several ways. There are
several types of inheriting a single gene disorder including but
not limited to autosomal dominant, autosomal recessive, X-linked
dominant, X-linked recessive, P-linked and mitochondrial
inheritance. Only one mutated copy of the gene will be necessary
for a person to be affected by an autosomal dominant disorder.
Examples of autosomal dominant type of disorder include but are not
limited to Huntington's disease, Neurofibromatosis 1, Marfan
Syndrome, Hereditary nonpolyposis colorectal cancer, and Hereditary
multiple exostoses. In autosomal recessive disorder, two copies of
the gene must be mutated for a person to be affected by an
autosomal recessive disorder. Examples of this type of disorder
include but are not limited to cystic fibrosis, sickle-cell disease
(also partial sickle-cell disease), Tay-Sachs disease, Niemann-Pick
disease, spinal muscular atrophy, and dry earwax. X-linked dominant
disorders are caused by mutations in genes on the X chromosome.
Only a few disorders have this inheritance pattern, with a prime
example being X-linked hypophosphatemic rickets. Males and females
are both affected in these disorders, with males typically being
more severely affected than females. Some X-linked dominant
conditions such as Rett syndrome, Incontinentia Pigmenti type 2 and
Aicardi Syndrome are usually fatal in males either in utero or
shortly after birth, and are therefore predominantly seen in
females. X-linked recessive disorders are also caused by mutations
in genes on the X chromosome. Examples of this type of disorder
include but are not limited to Hemophilia A, Duchenne muscular
dystrophy, red-green color blindness, muscular dystrophy and
Androgenetic alopecia. Y-linked disorders are caused by mutations
on the Y chromosome. Examples include but are not limited to Male
Infertility and hypertrichosis pinnae. Mitochondrial inheritance,
also known as maternal inheritance, applies to genes in
mitochondrial DNA. An example of this type of disorder is Leber's
Hereditary Optic Neuropathy.
[0068] Genetic disorders may also be complex, multifactorial or
polygenic, this means that they are likely associated with the
effects of multiple genes in combination with lifestyle and
environmental factors. Although complex disorders often cluster in
families, they do not have a clear-cut pattern of inheritance. This
makes it difficult to determine a person's risk of inheriting or
passing on these disorders. Complex disorders are also difficult to
study and treat because the specific factors that cause most of
these disorders have not yet been identified. Multifactoral or
polygenic disorders that can be diagnosed, characterized and/or
monitored using the algorithms and methods of the present invention
include but are not limited to heart disease, diabetes, asthma,
autism, autoimmune diseases such as multiple sclerosis, cancers,
ciliopathies, cleft palate, hypertension, inflammatory bowel
disease, mental retardation and obesity.
[0069] Other genetic disorders that can be diagnosed, characterized
and/or monitored using the algorithms and methods of the present
invention include but are not limited to 1p36 deletion syndrome,
21-hydroxylase deficiency, 22q11.2 deletion syndrome, 47,XYY
syndrome, 48, XXXX, 49, XXXXX, aceruloplasminemia, achondrogenesis,
type II, achondroplasia, acute intermittent porphyria,
adenylosuccinate lyase deficiency, Adrenoleukodystrophy, ALA
deficiency porphyria, ALA dehydratase deficiency, Alexander
disease, alkaptonuria, alpha-1 antitrypsin deficiency, Alstrom
syndrome, Alzheimer's disease (type 1, 2, 3, and 4), Amelogenesis
Imperfecta, amyotrophic lateral sclerosis, Amyotrophic lateral
sclerosis type 2, Amyotrophic lateral sclerosis type 4, amyotrophic
lateral sclerosis type 4, androgen insensitivity syndrome, Anemia,
Angelman syndrome, Apert syndrome, ataxia-telangiectasia,
Beare-Stevenson cutis gyrata syndrome, Benjamin syndrome, beta
thalassemia, biotinidase deficiency, Birt-Hogg-Dube syndrome,
bladder cancer, Bloom syndrome, Bone diseases, breast cancer,
CADASIL, Camptomelic dysplasia, Canavan disease, Cancer, Celiac
Disease, CGD Chronic Granulomatous Disorder, Charcot-Marie-Tooth
disease, Charcot-Marie-Tooth disease Type 1, Charcot-Marie-Tooth
disease Type 4, Charcot-Marie-Tooth disease, type 2,
Charcot-Marie-Tooth disease, type 4, Cockayne syndrome,
Coffin-Lowry syndrome, collagenopathy, types II and XI, Colorectal
Cancer, Congenital absence of the vas deferens, congenital
bilateral absence of vas deferens, congenital diabetes, congenital
erythropoietic porphyria, Congenital heart disease, congenital
hypothyroidism, Connective tissue disease, Cowden syndrome, Cri du
chat, Crohn's disease, fibrostenosing, Crouzon syndrome,
Crouzonodermoskeletal syndrome, cystic fibrosis, De Grouchy
Syndrome, Degenerative nerve diseases, Dent's disease,
developmental disabilities, DiGeorge syndrome, Distal spinal
muscular atrophy type V, Down syndrome, Dwarfism, Ehlers-Danlos
syndrome, Ehlers-Danlos syndrome arthrochalasia type, Ehlers-Danlos
syndrome classical type, Ehlers-Danlos syndrome dermatosparaxis
type, Ehlers-Danlos syndrome kyphoscoliosis type, vascular type,
erythropoietic protoporphyria, Fabry's disease, Facial injuries and
disorders, factor V Leiden thrombophilia, familial adenomatous
polyposis, familial dysautonomia, fanconi anemia, FG syndrome,
fragile X syndrome, Friedreich ataxia, Friedreich's ataxia, G6PD
deficiency, galactosemia, Gaucher's disease (type 1, 2, and 3),
Genetic brain disorders, Glycine encephalopathy, Haemochromatosis
type 2, Haemochromatosis type 4, Harlequin Ichthyosis, Head and
brain malformations, Hearing disorders and deafness, Hearing
problems in children, hemochromatosis (neonatal, type 2 and type
3), hemophilia, hepatoerythropoietic porphyria, hereditary
coproporphyria, Hereditary Multiple Exostoses, hereditary
neuropathy with liability to pressure palsies, hereditary
nonpolyposis colorectal cancer, homocystinuria, Huntington's
disease, Hutchinson Gilford Progeria Syndrome, hyperoxaluria,
primary, hyperphenylalaninemia, hypochondrogenesis,
hypochondroplasia, idic15, incontinentia pigmenti, Infantile
Gaucher disease, infantile-onset ascending hereditary spastic
paralysis, Infertility, Jackson-Weiss syndrome, Joubert syndrome,
Juvenile Primary Lateral Sclerosis, Kennedy disease, Klinefelter
syndrome, Kniest dysplasia, Krabbe disease, Learning disability,
Lesch-Nyhan syndrome, Leukodystrophies, Li-Fraumeni syndrome,
lipoprotein lipase deficiency, familial, Male genital disorders,
Marfan syndrome, McCune-Albright syndrome, McLeod syndrome,
Mediterranean fever, familial, MEDNIK, Menkes disease, Menkes
syndrome, Metabolic disorders, methemoglobinemia beta-globin type,
Methemoglobinemia congenital methaemoglobinaemia, methylmalonic
acidemia, Micro syndrome, Microcephaly, Movement disorders,
Mowat-Wilson syndrome, Mucopolysaccharidosis (MPS I), Muenke
syndrome, Muscular dystrophy, Muscular dystrophy, Duchenne and
Becker type, muscular dystrophy, Duchenne and Becker types,
myotonic dystrophy, Myotonic dystrophy type 1 and type 2, Neonatal
hemochromatosis, neurofibromatosis, neurofibromatosis 1,
neurofibromatosis 2, Neurofibromatosis type I, neurofibromatosis
type II, Neurologic diseases, Neuromuscular disorders, Niemann-Pick
disease, Nonketotic hyperglycinemia, nonsyndromic deafness,
Nonsyndromic deafness autosomal recessive, Noonan syndrome,
osteogenesis imperfecta (type I and type III),
otospondylomegaepiphyseal dysplasia, pantothenate kinase-associated
neurodegeneration, Patau Syndrome (Trisomy 13), Pendred syndrome,
Peutz-Jeghers syndrome, Pfeiffer syndrome, phenylketonuria,
porphyria, porphyria cutanea tarda, Prader-Willi syndrome, primary
pulmonary hypertension, prion disease, Progeria, propionic
acidemia, protein C deficiency, protein S deficiency,
pseudo-Gaucher disease, pseudoxanthoma elasticum, Retinal
disorders, retinoblastoma, retinoblastoma FA--Friedreich ataxia,
Rett syndrome, Rubinstein-Taybi syndrome, SADDAN, Sandhoff disease,
sensory and autonomic neuropathy type III, sickle cell anemia,
skeletal muscle regeneration, Skin pigmentation disorders, Smith
Lemli Opitz Syndrome, Speech and communication disorders, spinal
muscular atrophy, spinal-bulbar muscular atrophy, spinocerebellar
ataxia, spondyloepimetaphyseal dysplasia, Strudwick type,
spondyloepiphyseal dysplasia congenita, Stickler syndrome, Stickler
syndrome COL2A1, Tay-Sachs disease, tetrahydrobiopterin deficiency,
thanatophoric dysplasia, thiamine-responsive megaloblastic anemia
with diabetes mellitus and sensorineural deafness, Thyroid disease,
Tourette's Syndrome, Treacher Collins syndrome, triple X syndrome,
tuberous sclerosis, Turner syndrome, Usher syndrome, variegate
porphyria, von Hippel-Lindau disease, Waardenburg syndrome,
Weissenbacher-Zweymiiller syndrome, Wilson disease, Wolf-Hirschhorn
syndrome, Xeroderma Pigmentosum, X-linked severe combined
immunodeficiency, X-linked sideroblastic anemia, and X-linked
spinal-bulbar muscle atrophy.
[0070] In one embodiment, the subject methods and algorithm are
used to diagnose, characterize, and monitor thyroid cancer. Other
types of cancer that can be diagnosed, characterized and/or
monitored using the algorithms and methods of the present invention
include but are not limited to adrenal cortical cancer, anal
cancer, aplastic anemia, bile duct cancer, bladder cancer, bone
cancer, bone metastasis, central nervous system (CNS) cancers,
peripheral nervous system (PNS) cancers, breast cancer, Castleman's
disease, cervical cancer, childhood Non-Hodgkin's lymphoma, colon
and rectum cancer, endometrial cancer, esophagus cancer, Ewing's
family of tumors (e.g. Ewing's sarcoma), eye cancer, gallbladder
cancer, gastrointestinal carcinoid tumors, gastrointestinal stromal
tumors, gestational trophoblastic disease, hairy cell leukemia,
Hodgkin's disease, Kaposi's sarcoma, kidney cancer, laryngeal and
hypopharyngeal cancer, acute lymphocytic leukemia, acute myeloid
leukemia, children's leukemia, chronic lymphocytic leukemia,
chronic myeloid leukemia, liver cancer, lung cancer, lung carcinoid
tumors, Non-Hodgkin's lymphoma, male breast cancer, malignant
mesothelioma, multiple myeloma, myelodysplastic syndrome,
myeloproliferative disorders, nasal cavity and paranasal cancer,
nasopharyngeal cancer, neuroblastoma, oral cavity and oropharyngeal
cancer, osteosarcoma, ovarian cancer, pancreatic cancer, penile
cancer, pituitary tumor, prostate cancer, retinoblastoma,
rhabdomyosarcoma, salivary gland cancer, sarcoma (adult soft tissue
cancer), melanoma skin cancer, non-melanoma skin cancer, stomach
cancer, testicular cancer, thymus cancer, uterine cancer (e.g.
uterine sarcoma), vaginal cancer, vulvar cancer, and Waldenstrom's
macroglobulinemia.
[0071] In some embodiments, gene expression product markers of the
present invention may provide increased accuracy of genetic
disorder or cancer diagnosis through the use of multiple gene
expression product markers in low quantity and quality, and
statistical analysis using the algorithms of the present invention.
In particular, the present invention provides, but is not limited
to, methods of diagnosing, characterizing and classifying gene
expression profiles associated with thyroid cancers. The present
invention also provides algorithms for characterizing and
classifying thyroid tissue samples, and kits and compositions
useful for the application of said methods. The disclosure further
includes methods for running a molecular profiling business.
[0072] In one embodiment of the invention, markers and genes can be
identified to have differential expression in thyroid cancer
samples compared to thyroid benign samples. Illustrative examples
having a benign pathology include follicular adenoma, Hurthle cell
adenoma, lymphocytic thyroiditis, and nodular hyperplasia.
Illustrative examples having a malignant pathology include
follicular carcinoma, follicular variant of papillary thyroid
carcinoma, medullary carcinoma, and papillary thyroid
carcinoma.
[0073] Biological samples may be treated to extract nucleic acid
such as DNA or RNA. The nucleic acid may be contacted with an array
of probes of the present invention under conditions to allow
hybridization. The degree of hybridization may be assayed in a
quantitative matter using a number of methods known in the art. In
some cases, the degree of hybridization at a probe position may be
related to the intensity of signal provided by the assay, which
therefore is related to the amount of complementary nucleic acid
sequence present in the sample. Software can be used to extract,
normalize, summarize, and analyze array intensity data from probes
across the human genome or transcriptome including expressed genes,
exons, introns, and miRNAs. In some embodiments, the intensity of a
given probe in either the benign or malignant samples can be
compared against a reference set to determine whether differential
expression is occurring in a sample. An increase or decrease in
relative intensity at a marker position on an array corresponding
to an expressed sequence is indicative of an increase or decrease
respectively of expression of the corresponding expressed sequence.
Alternatively, a decrease in relative intensity may be indicative
of a mutation in the expressed sequence.
[0074] The resulting intensity values for each sample can be
analyzed using feature selection techniques including filter
techniques which assess the relevance of features by looking at the
intrinsic properties of the data, wrapper methods which embed the
model hypothesis within a feature subset search, and embedded
techniques in which the search for an optimal set of features is
built into a classifier algorithm.
[0075] Filter techniques useful in the methods of the present
invention include (1) parametric methods such as the use of two
sample t-tests, ANOVA analyses, Bayesian frameworks, and Gamma
distribution models (2) model free methods such as the use of
Wilcoxon rank sum tests, between-within class sum of squares tests,
rank products methods, random permutation methods, or TNoM which
involves setting a threshold point for fold-change differences in
expression between two datasets and then detecting the threshold
point in each gene that minimizes the number of missclassifications
(3) and multivariate methods such as bivariate methods, correlation
based feature selection methods (CFS), minimum redundancy maximum
relavance methods (MRMR), Markov blanket filter methods, and
uncorrelated shrunken centroid methods. Wrapper methods useful in
the methods of the present invention include sequential search
methods, genetic algorithms, and estimation of distribution
algorithms. Embedded methods useful in the methods of the present
invention include random forest algorithms, weight vector of
support vector machine algorithms, and weights of logistic
regression algorithms. Bioinformatics. 2007 Oct. 1; 23(19):2507-17
provides an overview of the relative merits of the filter
techniques provided above for the analysis of intensity data.
[0076] Selected features may then be classified using a classifier
algorithm. Illustrative algorithms include but are not limited to
methods that reduce the number of variables such as principal
component analysis algorithms, partial least squares methods, and
independent component analysis algorithms. Illustrative algorithms
further include but are not limited to methods that handle large
numbers of variables directly such as statistical methods and
methods based on machine learning techniques. Statistical methods
include penalized logistic regression, prediction analysis of
microarrays (PAM), methods based on shrunken centroids, support
vector machine analysis, and regularized linear discriminant
analysis. Machine learning techniques include bagging procedures,
boosting procedures, random forest algorithms, and combinations
thereof. Cancer Inform. 2008; 6: 77-97 provides an overview of the
classification techniques provided above for the analysis of
microarray intensity data.
[0077] The markers and genes of the present invention can be
utilized to characterize the cancerous or non-cancerous status of
cells or tissues. The present invention includes a method for
diagnosing benign tissues or cells from malignant tissues or cells
comprising determining the differential expression of a marker or
gene in a thyroid sample of a subject wherein said marker or gene
is a marker or gene listed in FIG. 2-6, 9-13, 16 or 17. The present
invention also includes methods for diagnosing medullary thyroid
carcinoma comprising determining the differential expression of a
marker or gene in a thyroid sample of a subject wherein said marker
or gene is a marker or gene listed in FIG. 10. The present
invention also includes methods for diagnosing thyroid pathology
subtypes comprising determining the differential expression of a
marker or gene in a thyroid sample of a subject wherein said marker
or gene is a marker or gene listed in FIG. 13. The present
invention also includes methods for diagnosing benign tissues or
cells from malignant tissues or cells comprising determining the
differential expression of an miRNA in a thyroid sample of a
subject wherein said miRNA is an miRNA listed in FIG. 16 or 17.
[0078] In accordance with the foregoing, the differential
expression of a gene, genes, markers, miRNAs, or a combination
thereof as disclosed herein may be determined using northern
blotting and employing the sequences as identified in herein to
develop probes for this purpose. Such probes may be composed of DNA
or RNA or synthetic nucleotides or a combination of the above and
may advantageously be comprised of a contiguous stretch of
nucleotide residues matching, or complementary to, a sequence as
identified in FIG. 2-6, 9-13, 16 or 17. Such probes will most
usefully comprise a contiguous stretch of at least 15-200 residues
or more including 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26,
27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43,
44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60,
61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 80, 85,
90, 95, 100, 110, 120, 130, 140, 150, 160, 175, or 200 nucleotides
or more, derived from one or more of the sequences as identified in
FIG. 2-6, 9-13, 16 or 17. Thus, where a single probe binds multiple
times to the transcriptome of a sample of cells that are cancerous,
or are suspected of being cancerous, or predisposed to become
cancerous, whereas binding of the same probe to a similar amount of
transcriptome derived from the genome of otherwise non-cancerous
cells of the same organ or tissue results in observably more or
less binding, this is indicative of differential expression of a
gene, multiple genes, markers, or miRNAs comprising, or
corresponding to, the sequences identified in FIG. 2-6, 9-13, 16 or
17 from which the probe sequenced was derived.
[0079] In one such embodiment, the elevated expression, as compared
to normal cells and/or tissues of the same organ, is determined by
measuring the relative rates of transcription of RNA, such as by
production of corresponding cDNAs and then analyzing the resulting
DNA using probes developed from the gene sequences as identified in
FIG. 2-6, 9-13, 16 or 17. Thus, the levels of cDNA produced by use
of reverse transcriptase with the full RNA complement of a cell
suspected of being cancerous produces a corresponding amount of
cDNA that can then be amplified using polymerase chain reaction, or
some other means, such as linear amplification, isothermal
amplification, NASB, or rolling circle amplification, to determine
the relative levels of resulting cDNA and, thereby, the relative
levels of gene expression.
[0080] Increased expression may also be determined using agents
that selectively bind to, and thereby detect, the presence of
expression products of the genes disclosed herein. For example, an
antibody, possibly a suitably labeled antibody, such as where the
antibody is bound to a fluorescent or radiolabel, may be generated
against one of the polypeptides comprising a sequence as identified
in FIG. 2-6, and 9-13, and said antibody will then react with,
binding either selectively or specifically, to a polypeptide
encoded by one of the genes that corresponds to a sequence
disclosed herein. Such antibody binding, especially relative extent
of such binding in samples derived from suspected cancerous, as
opposed to otherwise non-cancerous, cells and tissues, can then be
used as a measure of the extent of expression, or over-expression,
of the cancer-related genes identified herein. Thus, the genes
identified herein as being over-expressed in cancerous cells and
tissues may be over-expressed due to increased copy number, or due
to over-transcription, such as where the over-expression is due to
over-production of a transcription factor that activates the gene
and leads to repeated binding of RNA polymerase, thereby generating
large than normal amounts of RNA transcripts, which are
subsequently translated into polypeptides, such as the polypeptides
comprising amino acid sequences as identified in FIG. 2-6, and
9-13. Such analysis provides an additional means of ascertaining
the expression of the genes identified according to the invention
and thereby determining the presence of a cancerous state in a
sample derived from a patient to be tested, of the predisposition
to develop cancer at a subsequent time in said patient.
[0081] In employing the methods of the invention, it should be
borne in mind that gene or marker expression indicative of a
cancerous state need not be characteristic of every cell found to
be cancerous. Thus, the methods disclosed herein are useful for
detecting the presence of a cancerous condition within a tissue
where less than all cells exhibit the complete pattern of
over-expression. For example, a set of selected genes or markers,
comprising sequences homologous under stringent conditions, or at
least 90%, preferably 95%, identical to at least one of the
sequences as identified in FIG. 2-6, 9-13, 16 or 17, may be found,
using appropriate probes, either DNA or RNA, to be present in as
little as 60% of cells derived from a sample of tumorous, or
malignant, tissue while being absent from as much as 60% of cells
derived from corresponding non-cancerous, or otherwise normal,
tissue (and thus being present in as much as 40% of such normal
tissue cells). In one embodiment, such expression pattern is found
to be present in at least 70% of cells drawn from a cancerous
tissue and absent from at least 70% of a corresponding normal,
non-cancerous, tissue sample. In another embodiment, such
expression pattern is found to be present in at least 80% of cells
drawn from a cancerous tissue and absent from at least 80% of a
corresponding normal, non-cancerous, tissue sample. In another
embodiment, such expression pattern is found to be present in at
least 90% of cells drawn from a cancerous tissue and absent from at
least 90% of a corresponding normal, non-cancerous, tissue sample.
In another embodiment, such expression pattern is found to be
present in at least 100% of cells drawn from a cancerous tissue and
absent from at least 100% of a corresponding normal, non-cancerous,
tissue sample, although the latter embodiment may represent a rare
occurrence.
[0082] In some embodiments molecular profiling includes detection,
analysis, or quantification of nucleic acid (DNA, or RNA), protein,
or a combination thereof. The diseases or conditions to be
diagnosed by the methods of the present invention include for
example conditions of abnormal growth in one or more tissues of a
subject including but not limited to skin, heart, lung, kidney,
breast, pancreas, liver, muscle, smooth muscle, bladder, gall
bladder, colon, intestine, brain, esophagus, or prostate. In some
embodiments, the tissues analyzed by the methods of the present
invention include thyroid tissues.
[0083] In some embodiments, the diseases or conditions diagnosed by
the methods of the present invention include benign and malignant
hyperproliferative disorders including but not limited to cancers,
hyperplasias, or neoplasias. In some cases, the hyperproliferative
disorders diagnosed by the methods of the present invention include
but are not limited to breast cancer such as a ductal carcinoma in
duct tissue in a mammary gland, medullary carcinomas, colloid
carcinomas, tubular carcinomas, and inflammatory breast cancer;
ovarian cancer, including epithelial ovarian tumors such as
adenocarcinoma in the ovary and an adenocarcinoma that has migrated
from the ovary into the abdominal cavity; uterine cancer; cervical
cancer such as adenocarcinoma in the cervix epithelial including
squamous cell carcinoma and adenocarcinomas; prostate cancer, such
as a prostate cancer selected from the following: an adenocarcinoma
or an adenocarinoma that has migrated to the bone; pancreatic
cancer such as epitheliod carcinoma in the pancreatic duct tissue
and an adenocarcinoma in a pancreatic duct; bladder cancer such as
a transitional cell carcinoma in urinary bladder, urothelial
carcinomas (transitional cell carcinomas), tumors in the urothelial
cells that line the bladder, squamous cell carcinomas,
adenocarcinomas, and small cell cancers; leukemia such as acute
myeloid leukemia (AML), acute lymphocytic leukemia, chronic
lymphocytic leukemia, chronic myeloid leukemia, hairy cell
leukemia, myelodysplasia, myeloproliferative disorders, acute
myelogenous leukemia (AML), chronic myelogenous leukemia (CML),
mastocytosis, chronic lymphocytic leukemia (CLL), multiple myeloma
(MM), and myelodysplastic syndrome (MDS); bone cancer; lung cancer
such as non-small cell lung cancer (NSCLC), which is divided into
squamous cell carcinomas, adenocarcinomas, and large cell
undifferentiated carcinomas, and small cell lung cancer; skin
cancer such as basal cell carcinoma, melanoma, squamous cell
carcinoma and actinic keratosis, which is a skin condition that
sometimes develops into squamous cell carcinoma; eye
retinoblastoma; cutaneous or intraocular (eye) melanoma; primary
liver cancer (cancer that begins in the liver); kidney cancer;
AIDS-related lymphoma such as diffuse large B-cell lymphoma, B-cell
immunoblastic lymphoma and small non-cleaved cell lymphoma;
Kaposi's Sarcoma; viral-induced cancers including hepatitis B virus
(HBV), hepatitis C virus (HCV), and hepatocellular carcinoma; human
lymphotropic virus-type 1 (HTLV-1) and adult T-cell
leukemia/lymphoma; and human papilloma virus (HPV) and cervical
cancer; central nervous system cancers (CNS) such as primary brain
tumor, which includes gliomas (astrocytoma, anaplastic astrocytoma,
or glioblastoma multiforme), Oligodendroglioma, Ependymoma,
Meningioma, Lymphoma, Schwannoma, and Medulloblastoma; peripheral
nervous system (PNS) cancers such as acoustic neuromas and
malignant peripheral nerve sheath tumor (MPNST) including
neurofibromas and schwannomas, malignant fibrous cytoma, malignant
fibrous histiocytoma, malignant meningioma, malignant mesothelioma,
and malignant mixed Mullerian tumor; oral cavity and oropharyngeal
cancer such as, hypopharyngeal cancer, laryngeal cancer,
nasopharyngeal cancer, and oropharyngeal cancer; stomach cancer
such as lymphomas, gastric stromal tumors, and carcinoid tumors;
testicular cancer such as germ cell tumors (GCTs), which include
seminomas and nonseminomas, and gonadal stromal tumors, which
include Leydig cell tumors and Sertoli cell tumors; thymus cancer
such as to thymomas, thymic carcinomas, Hodgkin disease,
non-Hodgkin lymphomas carcinoids or carcinoid tumors; rectal
cancer; and colon cancer. In some cases, the diseases or conditions
diagnosed by the methods of the present invention include but are
not limited to thyroid disorders such as for example benign thyroid
disorders including but not limited to follicular adenomas, Hurthle
cell adenomas, lymphocytic throiditis, and thyroid hyperplasia. In
some cases, the diseases or conditions diagnosed by the methods of
the present invention include but are not limited to malignant
thyroid disorders such as for example follicular carcinomas,
follicular variant of papillary thyroid carcinomas, medullary
carcinomas, and papillary carcinomas. In some cases, the methods of
the present invention provide for a diagnosis of a tissue as
diseased or normal. In other cases, the methods of the present
invention provide for a diagnosis of normal, benign, or malignant.
In some cases, the methods of the present invention provide for a
diagnosis of benign/normal, or malignant. In some cases, the
methods of the present invention provide for a diagnosis of one or
more of the specific diseases or conditions provided herein.
II. Obtaining a Biological Sample
[0084] In some embodiments, the methods of the present invention
provide for obtaining a sample from a subject. As used herein, the
term subject refers to any animal (e.g. a mammal), including but
not limited to humans, non-human primates, rodents, dogs, pigs, and
the like. The methods of obtaining provided herein include methods
of biopsy including fine needle aspiration, core needle biopsy,
vacuum assisted biopsy, incisional biopsy, excisional biopsy, punch
biopsy, shave biopsy or skin biopsy. The sample may be obtained
from any of the tissues provided herein including but not limited
to skin, heart, lung, kidney, breast, pancreas, liver, muscle,
smooth muscle, bladder, gall bladder, colon, intestine, brain,
prostate, esophagus, or thyroid. Alternatively, the sample may be
obtained from any other source including but not limited to blood,
sweat, hair follicle, buccal tissue, tears, menses, feces, or
saliva. In some embodiments of the present invention, a medical
professional may obtain a biological sample for testing. In some
cases the medical professional may refer the subject to a testing
center or laboratory for submission of the biological sample. In
other cases, the subject may provide the sample. In some cases, a
molecular profiling business of the present invention may obtain
the sample.
[0085] The sample may be obtained by methods known in the art such
as the biopsy methods provided herein, swabbing, scraping,
phlebotomy, or any other methods known in the art. In some cases,
the sample may be obtained, stored, or transported using components
of a kit of the present invention. In some cases, multiple samples,
such as multiple thyroid samples may be obtained for diagnosis by
the methods of the present invention. In some cases, multiple
samples, such as one or more samples from one tissue type (e.g.
thyroid) and one or more samples from another tissue (e.g. buccal)
may be obtained for diagnosis by the methods of the present
invention. In some cases, multiple samples such as one or more
samples from one tissue type (e.g. thyroid) and one or more samples
from another tissue (e.g. buccal) may be obtained at the same or
different times. In some cases, the samples obtained at different
times are stored and/or analyzed by different methods. For example,
a sample may be obtained and analyzed by cytological analysis
(routine staining). In some cases, further sample may be obtained
from a subject based on the results of a cytological analysis. The
diagnosis of cancer may include an examination of a subject by a
physician, nurse or other medical professional. The examination may
be part of a routine examination, or the examination may be due to
a specific complaint including but not limited to one of the
following: pain, illness, anticipation of illness, presence of a
suspicious lump or mass, a disease, or a condition. The subject may
or may not be aware of the disease or condition. The medical
professional may obtain a biological sample for testing. In some
cases the medical professional may refer the subject to a testing
center or laboratory for submission of the biological sample.
[0086] In some cases, the subject may be referred to a specialist
such as an oncologist, surgeon, or endocrinologist for further
diagnosis. The specialist may likewise obtain a biological sample
for testing or refer the individual to a testing center or
laboratory for submission of the biological sample. In any case,
the biological sample may be obtained by a physician, nurse, or
other medical professional such as a medical technician,
endocrinologist, cytologist, phlebotomist, radiologist, or a
pulmonologist. The medical professional may indicate the
appropriate test or assay to perform on the sample, or the
molecular profiling business of the present disclosure may consult
on which assays or tests are most appropriately indicated. The
molecular profiling business may bill the individual or medical or
insurance provider thereof for consulting work, for sample
acquisition and or storage, for materials, or for all products and
services rendered.
[0087] In some embodiments of the present invention, a medical
professional need not be involved in the initial diagnosis or
sample acquisition. An individual may alternatively obtain a sample
through the use of an over the counter kit. Said kit may contain a
means for obtaining said sample as described herein, a means for
storing said sample for inspection, and instructions for proper use
of the kit. In some cases, molecular profiling services are
included in the price for purchase of the kit. In other cases, the
molecular profiling services are billed separately.
[0088] A sample suitable for use by the molecular profiling
business may be any material containing tissues, cells, nucleic
acids, genes, gene fragments, expression products, gene expression
products, or gene expression product fragments of an individual to
be tested. Methods for determining sample suitability and/or
adequacy are provided. A sample may include but is not limited to,
tissue, cells, or biological material from cells or derived from
cells of an individual. The sample may be a heterogeneous or
homogeneous population of cells or tissues. The biological sample
may be obtained using any method known to the art that can provide
a sample suitable for the analytical methods described herein.
[0089] The sample may be obtained by non-invasive methods including
but not limited to: scraping of the skin or cervix, swabbing of the
cheek, saliva collection, urine collection, feces collection,
collection of menses, tears, or semen. In other cases, the sample
is obtained by an invasive procedure including but not limited to:
biopsy, alveolar or pulmonary lavage, needle aspiration, or
phlebotomy. The method of biopsy may further include incisional
biopsy, excisional biopsy, punch biopsy, shave biopsy, or skin
biopsy. The method of needle aspiration may further include fine
needle aspiration, core needle biopsy, vacuum assisted biopsy, or
large core biopsy. In some embodiments, multiple samples may be
obtained by the methods herein to ensure a sufficient amount of
biological material. Methods of obtaining suitable samples of
thyroid are known in the art and are further described in the ATA
Guidelines for thryoid nodule management (Cooper et al. Thyroid
Vol. 16 No. 2 2006), herein incorporated by reference in its
entirety. Generic methods for obtaining biological samples are also
known in the art and further described in for example Ramzy,
Ibrahim Clinical Cytopathology and Aspiration Biopsy 2001 which is
herein incorporated by reference in its entirety. In one
embodiment, the sample is a fine needle aspirate of a thyroid
nodule or a suspected thyroid tumor. In some cases, the fine needle
aspirate sampling procedure may be guided by the use of an
ultrasound, X-ray, or other imaging device.
[0090] In some embodiments of the present invention, the molecular
profiling business may obtain the biological sample from a subject
directly, from a medical professional, from a third party, or from
a kit provided by the molecular profiling business or a third
party. In some cases, the biological sample may be obtained by the
molecular profiling business after the subject, a medical
professional, or a third party acquires and sends the biological
sample to the molecular profiling business. In some cases, the
molecular profiling business may provide suitable containers, and
excipients for storage and transport of the biological sample to
the molecular profiling business.
III. Storing the Sample
[0091] In some embodiments, the methods of the present invention
provide for storing the sample for a time such as seconds, minutes,
hours, days, weeks, months, years or longer after the sample is
obtained and before the sample is analyzed by one or more methods
of the invention. In some cases, the sample obtained from a subject
is subdivided prior to the step of storage or further analysis such
that different portions of the sample are subject to different
downstream methods or processes including but not limited to
storage, cytological analysis, adequacy tests, nucleic acid
extraction, molecular profiling or a combination thereof.
[0092] In some cases, a portion of the sample may be stored while
another portion of said sample is further manipulated. Such
manipulations may include but are not limited to molecular
profiling; cytological staining; nucleic acid (RNA or DNA)
extraction, detection, or quantification; gene expression product
(RNA or Protein) extraction, detection, or quantification;
fixation; and examination. The sample may be fixed prior to or
during storage by any method known to the art such as using
glutaraldehyde, formaldehyde, or methanol. In other cases, the
sample is obtained and stored and subdivided after the step of
storage for further analysis such that different portions of the
sample are subject to different downstream methods or processes
including but not limited to storage, cytological analysis,
adequacy tests, nucleic acid extraction, molecular profiling or a
combination thereof. In some cases, samples are obtained and
analyzed by for example cytological analysis, and the resulting
sample material is further analyzed by one or more molecular
profiling methods of the present invention. In such cases, the
samples may be stored between the steps of cytological analysis and
the steps of molecular profiling. Samples may be stored upon
acquisition to facilitate transport, or to wait for the results of
other analyses. In another embodiment, samples may be stored while
awaiting instructions from a physician or other medical
professional.
[0093] The acquired sample may be placed in a suitable medium,
excipient, solution, or container for short term or long term
storage. Said storage may require keeping the sample in a
refrigerated, or frozen environment. The sample may be quickly
frozen prior to storage in a frozen environment. The frozen sample
may be contacted with a suitable cryopreservation medium or
compound including but not limited to: glycerol, ethylene glycol,
sucrose, or glucose. A suitable medium, excipient, or solution may
include but is not limited to: hanks salt solution, saline,
cellular growth medium, an ammonium salt solution such as ammonium
sulphate or ammonium phosphate, or water. Suitable concentrations
of ammonium salts include solutions of about 0.1 g/ml, 0.2 g/ml,
0.3 g/ml, 0.4 g/ml, 0.5 g/ml, 0.6 g/ml, 0.7 g/ml, 0.8 g/ml, 0.9
g/ml, 1.0 g/ml, 1.1 g/ml, 1.2 g/ml, 1.3 g/ml, 1.4 g/ml, 1.5 g/ml,
1.6 g/ml, 1.7 g/ml, 1.8 g/ml, 1.9 g/ml, 2.0 g/ml, 2.2 g/ml, 2.3
g/ml, 2.5 g/ml or higher. The medium, excipient, or solution may or
may not be sterile.
[0094] The sample may be stored at room temperature or at reduced
temperatures such as cold temperatures (e.g. between about
20.degree. C. and about 0.degree. C.), or freezing temperatures,
including for example 0 C, -1 C, -2 C, -3 C, -4 C, -5 C, -6 C, -7
C, -8 C, -9 C, -10 C, -12 C, -14 C, -15 C, -16 C, -20 C, -22 C, -25
C, -28 C, -30 C, -35 C, -40 C, -45 C, -50 C, -60 C, -70 C, -80 C,
-100 C, -120 C, -140 C, -180 C, -190 C, or about -200 C. In some
cases, the samples may be stored in a refrigerator, on ice or a
frozen gel pack, in a freezer, in a cryogenic freezer, on dry ice,
in liquid nitrogen, or in a vapor phase equilibrated with liquid
nitrogen.
[0095] The medium, excipient, or solution may contain preservative
agents to maintain the sample in an adequate state for subsequent
diagnostics or manipulation, or to prevent coagulation. Said
preservatives may include citrate, ethylene diamine tetraacetic
acid, sodium azide, or thimersol. The medium, excipient or solution
may contain suitable buffers or salts such as Tris buffers or
phosphate buffers, sodium salts (e.g. NaCl), calcium salts,
magnesium salts, and the like. In some cases, the sample may be
stored in a commercial preparation suitable for storage of cells
for subsequent cytological analysis such as but not limited to
Cytyc ThinPrep, SurePath, or Monoprep.
[0096] The sample container may be any container suitable for
storage and or transport of the biological sample including but not
limited to: a cup, a cup with a lid, a tube, a sterile tube, a
vacuum tube, a syringe, a bottle, a microscope slide, or any other
suitable container. The container may or may not be sterile.
IV. Transportation of the Sample
[0097] The methods of the present invention provide for transport
of the sample. In some cases, the sample is transported from a
clinic, hospital, doctor's office, or other location to a second
location whereupon the sample may be stored and/or analyzed by for
example, cytological analysis or molecular profiling. In some
cases, the sample may be transported to a molecular profiling
company in order to perform the analyses described herein. In other
cases, the sample may be transported to a laboratory such as a
laboratory authorized or otherwise capable of performing the
methods of the present invention such as a Clinical Laboratory
Improvement Amendments (CLIA) laboratory. The sample may be
transported by the individual from whom the sample derives. Said
transportation by the individual may include the individual
appearing at a molecular profiling business or a designated sample
receiving point and providing a sample. Said providing of the
sample may involve any of the techniques of sample acquisition
described herein, or the sample may have already have been acquired
and stored in a suitable container as described herein. In other
cases the sample may be transported to a molecular profiling
business using a courier service, the postal service, a shipping
service, or any method capable of transporting the sample in a
suitable manner. In some cases, the sample may be provided to a
molecular profiling business by a third party testing laboratory
(e.g. a cytology lab). In other cases, the sample may be provided
to a molecular profiling business by the subject's primary care
physician, endocrinologist or other medical professional. The cost
of transport may be billed to the individual, medical provider, or
insurance provider. The molecular profiling business may begin
analysis of the sample immediately upon receipt, or may store the
sample in any manner described herein. The method of storage may or
may not be the same as chosen prior to receipt of the sample by the
molecular profiling business.
[0098] The sample may be transported in any medium or excipient
including any medium or excipient provided herein suitable for
storing the sample such as a cryopreservation medium or a liquid
based cytology preparation. In some cases, the sample may be
transported frozen or refrigerated such as at any of the suitable
sample storage temperatures provided herein.
[0099] Upon receipt of the sample by the molecular profiling
business, a representative or licensee thereof, a medical
professional, researcher, or a third party laboratory or testing
center (e.g. a cytology laboratory) the sample may be assayed using
a variety of routine analyses known to the art such as cytological
assays, and genomic analysis. Such tests may be indicative of
cancer, the type of cancer, any other disease or condition, the
presence of disease markers, or the absence of cancer, diseases,
conditions, or disease markers. The tests may take the form of
cytological examination including microscopic examination as
described below. The tests may involve the use of one or more
cytological stains. The biological material may be manipulated or
prepared for the test prior to administration of the test by any
suitable method known to the art for biological sample preparation.
The specific assay performed may be determined by the molecular
profiling company, the physician who ordered the test, or a third
party such as a consulting medical professional, cytology
laboratory, the subject from whom the sample derives, or an
insurance provider. The specific assay may be chosen based on the
likelihood of obtaining a definite diagnosis, the cost of the
assay, the speed of the assay, or the suitability of the assay to
the type of material provided.
V. Test for Adequacy
[0100] Subsequent to or during sample acquisition, including before
or after a step of storing the sample, the biological material may
be collected and assessed for adequacy, for example, to asses the
suitability of the sample for use in the methods and compositions
of the present invention. The assessment may be performed by the
individual who obtains the sample, the molecular profiling
business, the individual using a kit, or a third party such as a
cytological lab, pathologist, endocrinologist, or a researcher. The
sample may be determined to be adequate or inadequate for further
analysis due to many factors including but not limited to:
insufficient cells, insufficient genetic material, insufficient
protein, DNA, or RNA, inappropriate cells for the indicated test,
or inappropriate material for the indicated test, age of the
sample, manner in which the sample was obtained, or manner in which
the sample was stored or transported. Adequacy may be determined
using a variety of methods known in the art such as a cell staining
procedure, measurement of the number of cells or amount of tissue,
measurement of total protein, measurement of nucleic acid, visual
examination, microscopic examination, or temperature or pH
determination. In one embodiment, sample adequacy will be
determined from the results of performing a gene expression product
level analysis experiment. In another embodiment sample adequacy
will be determined by measuring the content of a marker of sample
adequacy. Such markers include elements such as iodine, calcium,
magnesium, phosphorous, carbon, nitrogen, sulfur, iron etc.;
proteins such as but not limited to thyroglobulin; cellular mass;
and cellular components such as protein, nucleic acid, lipid, or
carbohydrate.
[0101] In some cases, iodine may be measured by a chemical method
such as described in U.S. Pat. No. 3,645,691 which is incorporated
herein by reference in its entirety or other chemical methods known
in the art for measuring iodine content. Chemical methods for
iodine measurement include but are not limited to methods based on
the Sandell and Kolthoff reaction. Said reaction proceeds according
to the following equation:
2Ce.sup.4++As.sup.3+.fwdarw.2Ce.sup.3++As.sup.5+I.
[0102] Iodine has a catalytic effect upon the course of the
reaction, i.e., the more iodine present in the preparation to be
analyzed, the more rapidly the reaction proceeds. The speed of
reaction is proportional to the iodine concentration. In some
cases, this analytical method may carried out in the following
manner:
[0103] A predetermined amount of a solution of arsenous oxide
As.sub.2O.sub.3 in concentrated sulfuric or nitric acid is added to
the biological sample and the temperature of the mixture is
adjusted to reaction temperature, i.e., usually to a temperature
between 20.degree. C. and 60.degree. C. A predetermined amount of a
cerium (IV) sulfate solution in sulfuric or nitric acid is added
thereto. Thereupon, the mixture is allowed to react at the
predetermined temperature for a definite period of time. Said
reaction time is selected in accordance with the order of magnitude
of the amount of iodine to be determined and with the respective
selected reaction temperature. The reaction time is usually between
about 1 minute and about 40 minutes. Thereafter, the content of the
test solution of cerium (IV) ions is determined photometrically.
The lower the photometrically determined cerium (IV) ion
concentration is, the higher is the speed of reaction and,
consequently, the amount of catalytic agent, i.e., of iodine. In
this manner the iodine of the sample can directly and
quantitatively be determined.
[0104] In other cases, iodine content of a sample of thyroid tissue
may be measured by detecting a specific isotope of iodine such as
for example .sup.123I, .sup.124I, .sup.125I, and .sup.131I. In
still other cases, the marker may be another radioisotope such as
an isotope of carbon, nitrogen, sulfur, oxygen, iron, phosphorous,
or hydrogen. The radioisotope in some instances may be administered
prior to sample collection. Methods of radioisotope administration
suitable for adequacy testing are well known in the art and include
injection into a vein or artery, or by ingestion. A suitable period
of time between administration of the isotope and acquisition of
thyroid nodule sample so as to effect absorption of a portion of
the isotope into the thyroid tissue may include any period of time
between about a minute and a few days or about one week including
about 1 minute, 2 minutes, 5 minutes, 10 minutes, 15 minutes, 1/2
an hour, an hour, 8 hours, 12 hours, 24 hours, 48 hours, 72 hours,
or about one, one and a half, or two weeks, and may readily be
determined by one skilled in the art. Alternatively, samples may be
measured for natural levels of isotopes such as radioisotopes of
iodine, calcium, magnesium, carbon, nitrogen, sulfur, oxygen, iron,
phosphorous, or hydrogen.
[0105] (i) Cell and/or Tissue Content Adequacy Test
[0106] Methods for determining the amount of a tissue include but
are not limited to weighing the sample or measuring the volume of
sample. Methods for determining the amount of cells include but are
not limited to counting cells which may in some cases be performed
after dis-aggregation with for example an enzyme such as trypsin or
collagenase or by physical means such as using a tissue homogenizer
for example. Alternative methods for determining the amount of
cells recovered include but are not limited to quantification of
dyes that bind to cellular material, or measurement of the volume
of cell pellet obtained following centrifugation. Methods for
determining that an adequate number of a specific type of cell is
present include PCR, Q-PCR, RT-PCR, immuno-histochemical analysis,
cytological analysis, microscopic, and or visual analysis.
[0107] (ii) Nucleic Acid Content Adequacy Test
[0108] Samples may be analyzed by determining nucleic acid content
after extraction from the biological sample using a variety of
methods known to the art. In some cases, nucleic acids such as RNA
or mRNA is extracted from other nucleic acids prior to nucleic acid
content analysis. Nucleic acid content may be extracted, purified,
and measured by ultraviolet absorbance, including but not limited
to aborbance at 260 nanometers using a spectrophotometer. In other
cases nucleic acid content or adequacy may be measured by
fluorometer after contacting the sample with a stain. In still
other cases, nucleic acid content or adequacy may be measured after
electrophoresis, or using an instrument such as an agilent
bioanalyzer for example. It is understood that the methods of the
present invention are not limited to a specific method for
measuring nucleic acid content and or integrity.
[0109] In some embodiments, the RNA quantity or yield from a given
sample is measured shortly after purification using a NanoDrop
spectrophotometer in a range of nano- to micrograms. In some
embodiments, RNA quality is measured using an Agilent 2100
Bioanalyzer instrument, and is characterized by a calculated RNA
Integrity Number (RIN, 1-10). The NanoDrop is a cuvette-free
spectrophotometer. It uses 1 microleter to measure from 5 ng/.mu.l
to 3,000 ng/.mu.l of sample. The key features of NanoDrop include
low volume of sample and no cuvette; large dynamic range 5 ng/.mu.l
to 3,000 ng/.mu.l; and it allows quantitation of DNA, RNA and
proteins. NanoDrop.TM. 2000c allows for the analysis of 0.5
.mu.l-2.0 .mu.l samples, without the need for cuvettes or
capillaries.
[0110] RNA quality can be measured by a calculated RNA Integrity
Number (RIN). The RNA integrity number (RIN) is an algorithm for
assigning integrity values to RNA measurements. The integrity of
RNA is a major concern for gene expression studies and
traditionally has been evaluated using the 285 to 18S rRNA ratio, a
method that has been shown to be inconsistent. The RIN algorithm is
applied to electrophoretic RNA measurements and based on a
combination of different features that contribute information about
the RNA integrity to provide a more robust universal measure. In
some embodiments, RNA quality is measured using an Agilent 2100
Bioanalyzer instrument. The protocols for measuring RNA quality are
known and available commercially, for example, at Agilent website.
Briefly, in the first step, researchers deposit total RNA sample
into an RNA Nano LabChip. In the second step, the LabChip is
inserted into the Agilent bioanalyzer and let the analysis run,
generating a digital electropherogram. In the third step, the new
RIN algorithm then analyzes the entire electrophoretic trace of the
RNA sample, including the presence or absence of degradation
products, to determine sample integrity. Then, The algorithm
assigns a 1 to 10 RIN score, where level 10 RNA is completely
intact. Because interpretation of the electropherogram is automatic
and not subject to individual interpretation, universal and
unbiased comparison of samples is enabled and repeatability of
experiments is improved. The RIN algorithm was developed using
neural networks and adaptive learning in conjunction with a large
database of eukaryote total RNA samples, which were obtained mainly
from human, rat, and mouse tissues. Advantages of RIN include
obtain a numerical assessment of the integrity of RNA; directly
comparing RNA samples, e.g. before and after archival, compare
integrity of same tissue across different labs; and ensuring
repeatability of experiments, e.g. if RIN shows a given value and
is suitable for microarray experiments, then the RIN of the same
value can always be used for similar experiments given that the
same organism/tissue/extraction method is used (Schroeder A, et al.
BMC Molecular Biology 2006, 7:3 (2006)).
[0111] In some embodiments, RNA quality is measured on a scale of
RIN 1 to 10, 10 being highest quality. In one aspect, the present
invention provides a method of analyzing gene expression from a
sample with an RNA RIN value equal or less than 6.0. In some
embodiments, a sample containing RNA with an RIN number of 1.0,
2.0, 3.0, 4.0, 5.0 or 6.0 is analyzed for microarray gene
expression using the subject methods and algorithms of the present
invention. In some embodiments, the sample is a fine needle
aspirate of thyroid tissue. The sample can be degraded with an RIN
as low as 2.0.
[0112] Determination of gene expression in a given sample is a
complex, dynamic, and expensive process. RNA samples with
RIN.ltoreq.5.0 are typically not used for multi-gene microarray
analysis, and may instead be used only for single-gene RT-PCR
and/or TaqMan assays. This dichotomy in the usefulness of RNA
according to quality has thus far limited the usefulness of samples
and hampered research efforts. The present invention provides
methods via which low quality RNA can be used to obtain meaningful
multi-gene expression results from samples containing low
concentrations of RNA, for example, thyroid FNA samples.
[0113] In addition, samples having a low and/or un-measurable RNA
concentration by NanoDrop normally deemed inadequate for multi-gene
expression profiling can be measured and analyzed using the subject
methods and algorithms of the present invention. The most sensitive
and "state of the art" apparatus used to measure nucleic acid yield
in the laboratory today is the NanoDrop spectrophotometer. Like
many quantitative instruments of its kind, the accuracy of a
NanoDrop measurement decreases significantly with very low RNA
concentration. The minimum amount of RNA necessary for input into a
microarray experiment also limits the usefulness of a given sample.
In the present invention, a sample containing a very low amount of
nucleic acid can be estimated using a combination of the
measurements from both the NanoDrop and the Bioanalyzer
instruments, thereby optimizing the sample for multi-gene
expression assays and analysis.
[0114] (iii) Protein Content Adequacy Test
[0115] In some cases, protein content in the biological sample may
be measured using a variety of methods known to the art, including
but not limited to: ultraviolet absorbance at 280 nanometers, cell
staining as described herein, or protein staining with for example
coomassie blue, or bichichonic acid. In some cases, protein is
extracted from the biological sample prior to measurement of the
sample. In some cases, multiple tests for adequacy of the sample
may be performed in parallel, or one at a time. In some cases, the
sample may be divided into aliquots for the purpose of performing
multiple diagnostic tests prior to, during, or after assessing
adequacy. In some cases, the adequacy test is performed on a small
amount of the sample which may or may not be suitable for further
diagnostic testing. In other cases, the entire sample is assessed
for adequacy. In any case, the test for adequacy may be billed to
the subject, medical provider, insurance provider, or government
entity.
[0116] In some embodiments of the present invention, the sample may
be tested for adequacy soon or immediately after collection. In
some cases, when the sample adequacy test does not indicate a
sufficient amount sample or sample of sufficient quality,
additional samples may be taken.
VI. Analysis of Sample
[0117] In one aspect, the present invention provides methods for
performing microarray gene expression analysis with low quantity
and quality of polynucleotide, such as DNA or RNA. In some
embodiments, the present disclosure describes methods of
diagnosing, characterizing and/or monitoring a cancer by analyzing
gene expression with low quantity and quality of RNA. In one
embodiment, the cancer is thyroid cancer. Thyroid RNA can be
obtained from fine needle aspirates (FNA). In some embodiments,
gene expression profile is obtained from degraded samples with an
RNA RIN value of 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0 or
less. In particular embodiments, gene expression profile is
obtained from a sample with an RIN of equal or less than 6, i.e.
6.0, 5.0, 4.0, 3.0, 2.0, 1.0 or less. Provided by the present
invention are methods by which low quality RNA can be used to
obtain meaningful gene expression results from samples containing
low concentrations of nucleic acid, such as thyroid FNA
samples.
[0118] Another estimate of sample usefulness is RNA yield,
typically measured in nanogram to microgram amounts for gene
expression assays. The most sensitive and "state of the art"
apparatus used to measure nucleic acid yield in the laboratory
today is the NanoDrOp spectrophotometer. Like many quantitative
instruments of its kind, the accuracy of a NanoDrop measurement
decreases significantly with very low RNA concentration. The
minimum amount of RNA necessary for input into a microarray
experiment also limits the usefulness of a given sample. In some
aspects, the present invention solves the low RNA concentration
problem by estimating sample input using a combination of the
measurements from both the NanoDrop and the Bioanalyzer
instruments. Since the quality of data obtained from a gene
expression study is dependent on RNA quantity, meaningful gene
expression data can be generated from samples having a low or
un-measurable RNA concentration as measured by NanoDrop.
[0119] The subject methods and algorithms enable: 1) gene
expression analysis of samples containing low amount and/or low
quality of nucleic acid; 2) a significant reduction of false
positives and false negatives, 3) a determination of the underlying
genetic, metabolic, or signaling pathways responsible for the
resulting pathology, 4) the ability to assign a statistical
probability to the accuracy of the diagnosis of genetic disorders,
5) the ability to resolve ambiguous results, and 6) the ability to
distinguish between sub-types of cancer.
[0120] Cytological Analysis
[0121] Samples may be analyzed by cell staining combined with
microscopic examination of the cells in the biological sample. Cell
staining, or cytological examination, may be performed by a number
of methods and suitable reagents known to the art including but not
limited to: EA stains, hematoxylin stains, cytostain, papanicolaou
stain, eosin, nissl stain, toluidine blue, silver stain, azocarmine
stain, neutral red, or janus green. In some cases the cells are
fixed and/or permeablized with for example methanol, ethanol,
glutaraldehyde or formaldehyde prior to or during the staining
procedure. In some cases, the cells are not fixed. In some cases,
more than one stain is used in combination. In other cases no stain
is used at all. In some cases measurement of nucleic acid content
is performed using a staining procedure, for example with ethidium
bromide, hematoxylin, nissl stain or any nucleic acid stain known
to the art.
[0122] In some embodiments of the present invention, cells may be
smeared onto a slide by standard methods well known in the art for
cytological examination. In other cases, liquid based cytology
(LBC) methods may be utilized. In some cases, LBC methods provide
for an improved means of cytology slide preparation, more
homogenous samples, increased sensitivity and specificity, and
improved efficiency of handling of samples. In liquid based
cytology methods, biological samples are transferred from the
subject to a container or vial containing a liquid cytology
preparation solution such as for example Cytyc ThinPrep, SurePath,
or Monoprep or any other liquid based cytology preparation solution
known in the art. Additionally, the sample may be rinsed from the
collection device with liquid cytology preparation solution into
the container or vial to ensure substantially quantitative transfer
of the sample. The solution containing the biological sample in
liquid based cytology preparation solution may then be stored
and/or processed by a machine or by one skilled in the art to
produce a layer of cells on a glass slide. The sample may further
be stained and examined under the microscope in the same way as a
conventional cytological preparation.
[0123] In some embodiments of the present invention, samples may be
analyzed by immuno-histochemical staining. Immuno-histochemical
staining provides for the analysis of the presence, location, and
distribution of specific molecules or antigens by use of antibodies
in a biological sample (e.g. cells or tissues). Antigens may be
small molecules, proteins, peptides, nucleic acids or any other
molecule capable of being specifically recognized by an antibody.
Samples may be analyzed by immuno-histochemical methods with or
without a prior fixing and/or permeabilization step. In some cases,
the antigen of interest may be detected by contacting the sample
with an antibody specific for the antigen and then non-specific
binding may be removed by one or more washes. The specifically
bound antibodies may then be detected by an antibody detection
reagent such as for example a labeled secondary antibody, or a
labeled avidin/streptavidin. In some cases, the antigen specific
antibody may be labeled directly instead. Suitable labels for
immuno-histochemistry include but are not limited to fluorophores
such as fluoroscein and rhodamine, enzymes such as alkaline
phosphatase and horse radish peroxidase, and radionuclides such as
.sup.32P and .sup.125I. Gene product markers that may be detected
by immuno-histochemical staining include but are not limited to
Her2/Neu, Ras, Rho, EGFR, VEGFR, UbcH10, RET/PTC1, cytokeratin 20,
calcitonin, GAL-3, thyroid peroxidase, and thyroglobulin.
VII. Assay Results
[0124] The results of routine cytological or other assays may
indicate a sample as negative (cancer, disease or condition free),
ambiguous or suspicious (suggestive of the presence of a cancer,
disease or condition), diagnostic (positive diagnosis for a cancer,
disease or condition), or non diagnostic (providing inadequate
information concerning the presence or absence of cancer, disease,
or condition). The diagnostic results may be further classified as
malignant or benign. The diagnostic results may also provide a
score indicating for example, the severity or grade of a cancer, or
the likelihood of an accurate diagnosis, such as via a p-value, a
corrected p-value, or a statistical confidence indicator. In some
cases, the diagnostic results may be indicative of a particular
type of a cancer, disease, or condition, such as for example
follicular adenoma, Hurthle cell adenoma, lymphocytic thyroiditis,
hyperplasia, follicular carcinoma, follicular variant of papillary
thyroid carcinoma, papillary carcinoma, or any of the diseases or
conditions provided herein. In some cases, the diagnostic results
may be indicative of a particular stage of a cancer, disease, or
condition. The diagnostic results may inform a particular treatment
or therapeutic intervention for the type or stage of the specific
cancer disease or condition diagnosed. In some embodiments, the
results of the assays performed may be entered into a database. The
molecular profiling company may bill the individual, insurance
provider, medical provider, or government entity for one or more of
the following: assays performed, consulting services, reporting of
results, database access, or data analysis. In some cases all or
some steps other than molecular profiling are performed by a
cytological laboratory or a medical professional.
VIII. Molecular Profiling
[0125] Cytological assays mark the current diagnostic standard for
many types of suspected tumors including for example thyroid tumors
or nodules. In some embodiments of the present invention, samples
that assay as negative, indeterminate, diagnostic, or non
diagnostic may be subjected to subsequent assays to obtain more
information. In the present invention, these subsequent assays
comprise the steps of molecular profiling of genomic DNA, RNA, mRNA
expression product levels, miRNA levels, gene expression product
levels or gene expression product alternative splicing. In some
embodiments of the present invention, molecular profiling means the
determination of the number (e.g. copy number) and/or type of
genomic DNA in a biological sample. In some cases, the number
and/or type may further be compared to a control sample or a sample
considered normal. In some embodiment, genomic DNA can be analyzed
for copy number variation, such as an increase (amplification) or
decrease in copy number, or variants, such as insertions,
deletions, truncations and the like. Molecular profiling may be
performed on the same sample, a portion of the same sample, or a
new sample may be acquired using any of the methods described
herein. The molecular profiling company may request additional
sample by directly contacting the individual or through an
intermediary such as a physician, third party testing center or
laboratory, or a medical professional. In some cases, samples are
assayed using methods and compositions of the molecular profiling
business in combination with some or all cytological staining or
other diagnostic methods. In other cases, samples are directly
assayed using the methods and compositions of the molecular
profiling business without the previous use of routine cytological
staining or other diagnostic methods. In some cases the results of
molecular profiling alone or in combination with cytology or other
assays may enable those skilled in the art to diagnose or suggest
treatment for the subject. In some cases, molecular profiling may
be used alone or in combination with cytology to monitor tumors or
suspected tumors over time for malignant changes.
[0126] The molecular profiling methods of the present invention
provide for extracting and analyzing protein or nucleic acid (RNA
or DNA) from one or more biological samples from a subject. In some
cases, nucleic acid is extracted from the entire sample obtained.
In other cases, nucleic acid is extracted from a portion of the
sample obtained. In some cases, the portion of the sample not
subjected to nucleic acid extraction may be analyzed by cytological
examination or immuno-histochemistry. Methods for RNA or DNA
extraction from biological samples are well known in the art and
include for example the use of a commercial kit, such as the Qiagen
DNeasy Blood and Tissue Kit, or the Qiagen EZ1 RNA Universal Tissue
Kit.
[0127] (i) Tissue-Type Fingerprinting
[0128] In many cases, biological samples such as those provided by
the methods of the present invention of may contain several cell
types or tissues, including but not limited to thyroid follicular
cells, thyroid medullary cells, blood cells (RBCs, WBCs,
platelets), smooth muscle cells, ducts, duct cells, basement
membrane, lumen, lobules, fatty tissue, skin cells, epithelial
cells, and infiltrating macrophages and lymphocytes. In the case of
thyroid samples, diagnostic classification of the biological
samples may involve for example primarily follicular cells (for
cancers derived from the follicular cell such as papillary
carcinoma, follicular carcinoma, and anaplastic thyroid carcinoma)
and medullary cells (for medullary cancer). The diagnosis of
indeterminate biological samples from thyroid biopsies in some
cases concerns the distinction of follicular adenoma vs. follicular
carcinoma. The molecular profiling signal of a follicular cell for
example may thus be diluted out and possibly confounded by other
cell types present in the sample. Similarly diagnosis of biological
samples from other tissues or organs often involves diagnosing one
or more cell types among the many that may be present in the
sample.
[0129] In some embodiments, the methods of the present invention
provide for an upfront method of determining the cellular make-up
of a particular biological sample so that the resulting molecular
profiling signatures can be calibrated against the dilution effect
due to the presence of other cell and/or tissue types. In one
aspect, this upfront method is an algorithm that uses a combination
of known cell and/or tissue specific gene expression patterns as an
upfront mini-classifier for each component of the sample. This
algorithm utilizes this molecular fingerprint to pre-classify the
samples according to their composition and then apply a
correction/normalization factor. This data may in some cases then
feed in to a final classification algorithm which would incorporate
that information to aid in the final diagnosis.
[0130] (ii) Genomic Analysis
[0131] In some embodiments, genomic sequence analysis, or
genotyping, may be performed on the sample. This genotyping may
take the form of mutational analysis such as single nucleotide
polymorphism (SNP) analysis, insertion deletion polymorphism
(InDel) analysis, variable number of tandem repeat (VNTR) analysis,
copy number variation (CNV) analysis or partial or whole genome
sequencing. Methods for performing genomic analyses are known to
the art and may include high throughput sequencing such as but not
limited to those methods described in U.S. Pat. Nos. 7,335,762;
7,323,305; 7,264,929; 7,244,559; 7,211,390; 7,361,488; 7,300,788;
and 7,280,922. Methods for performing genomic analyses may also
include microarray methods as described hereinafter. In some cases,
genomic analysis may be performed in combination with any of the
other methods herein. For example, a sample may be obtained, tested
for adequacy, and divided into aliquots. One or more aliquots may
then be used for cytological analysis of the present invention, one
or more may be used for RNA expression profiling methods of the
present invention, and one or more can be used for genomic
analysis. It is further understood the present invention
anticipates that one skilled in the art may wish to perform other
analyses on the biological sample that are not explicitly provided
herein.
[0132] (iii) Expression Product Profiling
[0133] Gene expression profiling is the measurement of the activity
(the expression) of thousands of genes at once, to create a global
picture of cellular function. These profiles can, for example,
distinguish between cells that are actively dividing, or show how
the cells react to a particular treatment. Many experiments of this
sort measure an entire genome simultaneously, that is, every gene
present in a particular cell. Microarray technology measures the
relative activity of previously identified target genes. Sequence
based techniques, like serial analysis of gene expression (SAGE,
SuperSAGE) are also used for gene expression profiling. SuperSAGE
is especially accurate and can measure any active gene, not just a
predefined set. In an RNA, mRNA or gene expression profiling
microarray, the expression levels of thousands of genes are
simultaneously monitored to study the effects of certain
treatments, diseases, and developmental stages on gene expression.
For example, microarray-based gene expression profiling can be used
to characterize gene signatures of a genetic disorder disclosed
herein, or different cancer types, subtypes of a cancer, and/or
cancer stages.
[0134] Expression profiling experiments often involve measuring the
relative amount of gene expression products, such as mRNA,
expressed in two or more experimental conditions. This is because
altered levels of a specific sequence of a gene expression product
suggest a changed need for the protein coded for by the gene
expression product, perhaps indicating a homeostatic response or a
pathological condition. For example, if breast cancer cells express
higher levels of mRNA associated with a particular transmembrane
receptor than normal cells do, it might be that this receptor plays
a role in breast cancer. One aspect of the present invention
encompasses gene expression profiling as part of an important
diagnostic test for genetic disorders and cancers, particularly,
thyroid cancer.
[0135] In some embodiments, RNA samples with RIN.ltoreq.5.0 are
typically not used for multi-gene microarray analysis, and may
instead be used only for single-gene RT-PCR and/or TaqMan assays.
Microarray, RT-PCR and TaqMan assays are standard molecular
techniques well known in the relevant art. TaqMan probe-based
assays are widely used in real-time PCR including gene expression
assays, DNA quantification and SNP genotyping.
[0136] In one embodiment, gene expression products related to
cancer that are known to the art are profiled. Such gene expression
products have been described and include but are not limited to the
gene expression products detailed in U.S. Pat. Nos. 7,358,061;
7,319,011; 5,965,360; 6,436,642; and US patent applications
2003/0186248, 2005/0042222, 2003/0190602, 2005/0048533,
2005/0266443, 2006/0035244, 2006/083744, 2006/0088851,
2006/0105360, 2006/0127907, 2007/0020657, 2007/0037186,
2007/0065833, 2007/0161004, 2007/0238119, and 2008/0044824.
[0137] It is further anticipated that other gene expression
products related to cancer may become known, and that the methods
and compositions described herein may include such newly discovered
gene expression products.
[0138] In some embodiments of the present invention gene expression
products are analyzed alternatively or additionally for
characteristics other than expression level. For example, gene
products may be analyzed for alternative splicing. Alternative
splicing, also referred to as alternative exon usage, is the RNA
splicing variation mechanism wherein the exons of a primary gene
transcript, the pre-mRNA, are separated and reconnected (i.e.
spliced) so as to produce alternative mRNA molecules from the same
gene. In some cases, these linear combinations then undergo the
process of translation where a specific and unique sequence of
amino acids is specified by each of the alternative mRNA molecules
from the same gene resulting in protein isoforms. Alternative
splicing may include incorporating different exons or different
sets of exons, retaining certain introns, or using utilizing
alternate splice donor and acceptor sites.
[0139] In some cases, markers or sets of markers may be identified
that exhibit alternative splicing that is diagnostic for benign,
malignant or normal samples. Additionally, alternative splicing
markers may further provide a diagnosis for the specific type of
thyroid cancer (e.g. papillary, follicular, medullary, or
anaplastic). Alternative splicing markers diagnostic for malignancy
known to the art include those listed in U.S. Pat. No.
6,436,642.
[0140] In some cases expression of RNA expression products that do
not encode for proteins such as miRNAs, and siRNAs may be assayed
by the methods of the present invention. Differential expression of
these RNA expression products may be indicative of benign,
malignant or normal samples. Differential expression of these RNA
expression products may further be indicative of the subtype of the
benign sample (e.g. FA, NHP, LCT, BN, CN, HA) or malignant sample
(e.g. FC, PTC, FVPTC, ATC, MTC). In some cases, differential
expression of miRNAs, siRNAs, alternative splice RNA isoforms,
mRNAs or any combination thereof may be assayed by the methods of
the present invention.
[0141] In some embodiments, the current invention provides 16
panels of biomarkers, each panel being required to characterize,
rule out, and diagnose pathology within the thyroid. The sixteen
panels are:
[0142] 1 Normal Thyroid (NML)
[0143] 2 Lymphocytic, Autoimmune Thyroiditis (LCT)
[0144] 3 Nodular Hyperplasia (NHP)
[0145] 4 Follicular Thyroid Adenoma (FA)
[0146] 5 Hurthle Cell Thyroid Adenoma (HC)
[0147] 6 Parathyroid (non thyroid tissue)
[0148] 7 Anaplastic Thyroid Carcinoma (ATC)
[0149] 8 Follicular Thyroid Carcinoma (FC)
[0150] 9 Hurthle Cell Thyroid Carcinoma (HC)
[0151] 10 Papillary Thyroid Carcinoma (PTC)
[0152] 11 Follicular Variant of Papillary Carcinoma (FVPTC)
[0153] 12 Medullary Thyroid Carcinoma (MTC)
[0154] 13 Renal Carcinoma metastasis to the Thyroid
[0155] 14 Melanoma metastasis to the Thyroid
[0156] 15 B cell Lymphoma metastasis to the Thyroid
[0157] 16 Breast Carcinoma metastasis to the Thyroid
[0158] Each panel includes a set of biomarkers required to
characterize, rule out, and diagnose a given pathology within the
thyroid. Panels 1-6 describe benign pathology. Panels 7-16 describe
malignant pathology.
[0159] The biological nature of the thyroid and each pathology
found within it, suggests that there is redundancy between the
plurality of biomarkers in one panel versus the plurality of
biomarkers in another panel. Mirroring each pathology subtype, each
diagnostic panel is heterogeneous and semi-redundant with the
biomarkers in another panel. Heterogeneity and redundancy reflect
the biology of the tissues sampled in a given FNA and the
differences in gene expression that characterize each pathology
subtype from one another.
[0160] In one aspect, the diagnostic value of the present invention
lies in the comparison of i) one or more markers in one panel,
versus ii) one or more markers in each additional panel. The
utility of the invention is its higher diagnostic accuracy in FNA
than presently possible by any other means.
[0161] In some embodiments, the biomarkers within each panel are
interchangeable (modular). The plurality of biomarkers in all
panels can be substituted, increased, reduced, or improved to
accommodate the definition of new pathologic subtypes (e.g. new
case reports of metastasis to the thyroid from other organs). The
current invention describes the plurality of markers that define
each of sixteen heterogeneous, semi-redundant, and distinct
pathologies found in the thyroid. All sixteen panels are required
to arrive at an accurate diagnosis, and any given panel alone does
not have sufficient power to make a true diagnostic determination.
In some embodiments, the biomarkers in each panel are interchanged
with a suitable combination of biomarkers, such that the plurality
of biomarkers in each panel still defines a given pathology subtype
within the context of examining the plurality of biomarkers that
define all other pathology subtypes.
[0162] Methods and compositions of the invention can have genes
selected from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or
16 or more biomarker panels and can have from 1, 2; 3, 4, 5, 6, 7,
8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more gene expression
products from each biomarker panel, in any combination. In some
embodiments, the set of genes combined give a specificity or
sensitivity of greater than 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%,
90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 99.5%, or a
positive predictive value or negative predictive value of at least
95%, 95.5%, 96%, 96.5%, 97%, 97.5%, 98%, 98.5%, 99%, 99.5% or
more.
[0163] (1) In Vitro Methods of Determining Expression Product
Levels
[0164] The general methods for determining gene expression product
levels are known to the art and may include but are not limited to
one or more of the following: additional cytological assays, assays
for specific proteins or enzyme activities, assays for specific
expression products including protein or RNA or specific RNA splice
variants, in situ hybridization, whole or partial genome expression
analysis, microarray hybridization assays, SAGE, enzyme linked
immuno-absorbance assays, mass-spectrometry, immuno-histochemistry,
or blotting. Gene expression product levels may be normalized to an
internal standard such as total mRNA or the expression level of a
particular gene including but not limited to glyceraldehyde 3
phosphate dehydrogenase, or tublin.
[0165] In some-embodiments of the present invention, gene
expression product markers and alternative splicing markers may be
determined by microarray analysis using, for example, Affymetrix
arrays, cDNA microarrays, oligonucleotide microarrays, spotted
microarrays, or other microarray products from Biorad, Agilent, or
Eppendorf. Microarrays provide particular advantages because they
may contain a large number of genes or alternative splice variants
that may be assayed in a single experiment. In some cases, the
microarray device may contain the entire human genome or
transcriptome or a substantial fraction thereof allowing a
comprehensive evaluation of gene expression patterns, genomic
sequence, or alternative splicing. Markers may be found using
standard molecular biology and microarray analysis techniques as
described in Sambrook Molecular Cloning a Laboratory Manual 2001
and Baldi, P., and Hatfield, W. G., DNA Microarrays and Gene
Expression 2002.
[0166] Microarray analysis begins with extracting and purifying
nucleic acid from a biological sample, (e.g. a biopsy or fine
needle aspirate) using methods known to the art. For expression and
alternative splicing analysis it may be advantageous to extract
and/or purify RNA from DNA. It may further be advantageous to
extract and/or purify mRNA from other forms of RNA such as tRNA and
rRNA.
[0167] Purified nucleic acid may further be labeled with a
fluorescent, radionuclide, or chemical label such as biotin or
digoxin for example by reverse transcription, PCR, ligation,
chemical reaction or other techniques. The labeling can be direct
or indirect which may further require a coupling stage. The
coupling stage can occur before hybridization, for example, using
aminoallyl-UTP and NHS amino-reactive dyes (like cyanine dyes) or
after, for example, using biotin and labelled streptavidin. The
modified nucleotides (e.g. at a 1 aaUTP: 4 UP ratio) are added
enzymatically at a lower rate compared to normal nucleotides,
typically resulting in 1 every 60 bases (measured with a
spectrophotometer). The aaDNA may then be purified with, for
example, a column or a diafiltration device. The aminoallyl group
is an amine group on a long linker attached to the nucleobase,
which reacts with a reactive label (e.g. a fluorescent dye).
[0168] The labeled samples may then be mixed with a hybridization
solution which may contain SDS, SSC, dextran sulfate, a blocking
agent (such as COT1 DNA, salmon sperm DNA, calf thymum DNA, PolyA
or PolyT), Denhardt's solution, formamine, or a combination
thereof.
[0169] A hybridization probe is a fragment of DNA or RNA of
variable length, which is used to detect in DNA or RNA samples the
presence of nucleotide sequences (the DNA target) that are
complementary to the sequence in the probe. The probe thereby
hybridizes to single-stranded nucleic acid (DNA or RNA) whose base
sequence allows probe-target base pairing due to complementarity
between the probe and target. The labeled probe is first denatured
(by heating or under alkaline conditions) into single DNA strands
and then hybridized to the target DNA.
[0170] To detect hybridization of the probe to its target sequence,
the probe is tagged (of labeled) with a molecular marker; commonly
used markers are .sup.32P or Digoxigenin, which is non-radioactive
antibody-based marker. DNA sequences or RNA transcripts that have
moderate to high sequence similarity to the probe are then detected
by visualizing the hybridized probe via autoradiography or other
imaging techniques. Detection of sequences with moderate or high
similarity depends on how stringent the hybridization conditions
were applied--high stringency, such as high hybridization
temperature and low salt in hybridization buffers, permits only
hybridization between nucleic acid sequences that are highly
similar, whereas low stringency, such as lower temperature and high
salt, allows hybridization when the sequences are less similar.
Hybridization probes used in DNA microarrays refer to DNA
covalently attached to an inert surface, such as coated glass
slides or gene chips, and to which a mobile cDNA target is
hybridized.
[0171] This mix may then be denatured by heat or chemical means and
added to a port in a microarray. The holes may then be sealed and
the microarray hybridized, for example, in a hybridization oven,
where the microarray is mixed by rotation, or in a mixer. After an
overnight hybridization, non specific binding may be washed off
(e.g. with SDS and SSC). The microarray may then be dried and
scanned in a special machine where a laser excites the dye and a
detector measures its emission. The image may be overlaid with a
template grid and the intensities of the features (several pixels
make a feature) may be quantified.
[0172] Various kits can be used for the amplification of nucleic
acid and probe generation of the subject methods. Examples of kit
that can be used in the present invention include but are not
limited to Nugen WT-Ovation FFPE kit, cDNA amplification kit with
Nugen Exon Module and Frag/Label module. The NuGEN WT-Ovation.TM.
FFPE System V2 is a whole transcriptome amplification system that
enables conducting global gene expression analysis on the vast
archives of small and degraded RNA derived from FFPE samples. The
system is comprised of reagents and a protocol required for
amplification of as little as 50 ng of total FFPE RNA. The protocol
can be used for qPCR, sample archiving, fragmentation, and
labeling. The amplified cDNA can be fragmented and labeled in less
than two hours for GeneChip.RTM. 3' expression array analysis using
NuGEN's FL-Ovation.TM. cDNA Biotin Module V2. For analysis using
Affymetrix GeneChip.RTM. Exon and Gene ST arrays, the amplified
cDNA can be used with the WT-Ovation Exon Module, then fragmented
and labeled using the FL-Ovation.TM. cDNA Biotin Module V2. For
analysis on Agilent arrays, the amplified cDNA can be fragmented
and labeled using NuGEN's FL-Ovation.TM. cDNA Fluorescent Module.
More information on Nugen WT-Ovation FFPE kit can be obtained at
http://www.nugeninc.com/nugen/index.cfm/products/amplification-systems/wt-
-ovation-ffpe/.
[0173] In some embodiments, Ambion WT-expression kit can be used.
Ambion WT-expression kit allows amplification of total RNA directly
without a separate ribosomal RNA (rRNA) depletion step. With the
Ambion.RTM. WT Expression Kit, samples as small as 50 ng of total
RNA can be analyzed on Affymetrix.RTM. GeneChip.RTM. Human, Mouse,
and Rat Exon and Gene 1.0 ST Arrays. In addition to the lower input
RNA requirement and high concordance between the Affymetrix.RTM.
method and TaqMan.RTM. real-time PCR data, the Ambion.RTM. WT
Expression Kit provides a significant increase in sensitivity. For
example, a greater number of probe sets detected above background
can be obtained at the exon level with the Ambion.RTM. WT
Expression Kit as a result of an increased signal-to-noise ratio.
Ambion WT-expression kit may be used in combination with additional
Affymetrix labeling kit.
[0174] In some embodiments, AmpTec Trinucleotide Nano mRNA
Amplification kit (6299-A15) can be used in the subject methods.
The ExpressArt.RTM. TRinucleotide mRNA amplification Nano kit is
suitable for a wide range, from 1 ng to 700 ng of input total RNA.
According to the amount of input total RNA and the required yields
of aRNA, it can be used for 1-round (input>300 ng total RNA) or
2-rounds (minimal input amount 1 ng total RNA), with aRNA yields in
the range of >10 .mu.g. AmpTec's proprietary TRinucleotide
priming technology results in preferential amplification of mRNAs
(independent of the universal eukaryotic 3'-poly(A)-sequence),
combined with selection against rRNAs. More information on AmpTec
Trinucleotide Nano mRNA Amplification kit can be obtained at
http://www.amp-tec.com/products.htm. This kit can be used in
combination with cDNA conversion kit and Affymetrix labeling
kit.
[0175] The raw data may then be normalized, for example, by
subtracting the background intensity and then dividing the
intensities making either the total intensity of the features on
each channel equal or the intensities of a reference gene and then
the t-value for all the intensities may be calculated. More
sophisticated methods, include z-ratio, loess and lowess regression
and RMA (robust multichip analysis) for Affymetrix chips.
[0176] (2) In Vivo Methods of Determining Gene Expression Product
Levels
[0177] It is further anticipated that the methods and compositions
of the present invention may be used to determine gene expression
product levels in an individual without first obtaining a sample.
For example, gene expression product levels may be determined in
vivo, that is in the individual. Methods for determining gene
expression product levels in vivo are known to the art and include
imaging techniques such as CAT, MM; NMR; PET; and optical,
fluorescence, or biophotonic imaging of protein or RNA levels using
antibodies or molecular beacons. Such methods are described in US
2008/0044824, US 2008/0131892, herein incorporated by reference.
Additional methods for in vivo molecular profiling are contemplated
to be within the scope of the present invention.
[0178] In some embodiments of the present invention, molecular
profiling includes the step of binding the sample or a portion of
the sample to one or more probes of the present invention. Suitable
probes bind to components of the sample, i.e. gene products, that
are to be measured and include but are not limited to antibodies or
antibody fragments, aptamers, nucleic acids, and oligonucleotides.
The binding of the sample to the probes of the present invention
represents a transformation of matter from sample to sample bound
to one or more probes. The method of diagnosing cancer based on
molecular profiling further comprises the steps of detecting gene
expression products (i.e. mRNA or protein) and levels of the
sample, comparing it to an amount in a normal control sample to
determine the differential gene expression product level between
the sample and the control; and classifying the test sample by
inputting one or more differential gene expression product levels
to a trained algorithm of the present invention; validating the
sample classification using the selection and classification
algorithms of the present invention; and identifying the sample as
positive for a genetic disorder or a type of cancer.
[0179] (i) Comparison of Sample to Normal
[0180] The results of the molecular profiling performed on the
sample provided by the individual (test sample) may be compared to
a biological sample that is known or suspected to be normal. A
normal sample is that which is or is expected to be free of any
cancer, disease, or condition, or a sample that would test negative
for any cancer disease or condition in the molecular profiling
assay. The normal sample may be from a different individual from
the individual being tested, or from the same individual. In some
cases, the normal sample is a sample obtained from a buccal swab of
an individual such as the individual being tested for example. The
normal sample may be assayed at the same time, or at a different
time from the test sample.
[0181] The results of an assay on the test sample may be compared
to the results of the same assay on a normal sample. In some cases
the results of the assay on the normal sample are from a database,
or a reference. In some cases, the results of the assay on the
normal sample are a known or generally accepted value by those
skilled in the art. In some cases the comparison is qualitative. In
other cases the comparison is quantitative. In some cases,
qualitative or quantitative comparisons may involve but are not
limited to one or more of the following: comparing fluorescence
values, spot intensities, absorbance values, chemiluminescent
signals, histograms, critical threshold values, statistical
significance values, gene product expression levels, gene product
expression level changes, alternative exon usage, changes in
alternative exon usage, protein levels, DNA polymorphisms, coy
number variations, indications of the presence or absence of one or
more DNA markers or regions, or nucleic acid sequences.
[0182] (ii) Evaluation of Results
[0183] In some embodiments, the molecular profiling results are
evaluated using methods known to the art for correlating gene
product expression levels or alternative exon usage with specific
phenotypes such as malignancy, the type of malignancy (e.g.
follicular carcinoma), benignancy, or normalcy (e.g. disease or
condition free). In some cases, a specified statistical confidence
level may be determined in order to provide a diagnostic confidence
level. For example, it may be determined that a confidence level of
greater than 90% may be a useful predictor of malignancy, type of
malignancy, or benignancy. In other embodiments, more or less
stringent confidence levels may be chosen. For example, a
confidence level of approximately 70%, 75%, 80%, 85%, 90%, 95%,
97.5%, 99%, 99.5%, or 99.9% may be chosen as a useful phenotypic
predictor. The confidence level provided may in some cases be
related to the quality of the sample, the quality of the data, the
quality of the analysis, the specific methods used, and the number
of gene expression products analyzed. The specified confidence
level for providing a diagnosis may be chosen on the basis of the
expected number of false positives or false negatives and/or cost.
Methods for choosing parameters for achieving a specified
confidence level or for identifying markers with diagnostic power
include but are not limited to Receiver Operator Curve analysis
(ROC), binormal ROC, principal component analysis, partial least
squares analysis, singular value decomposition, least absolute
shrinkage and selection operator analysis, least angle regression,
and the threshold gradient directed regularization method.
[0184] (iii) Data Analysis
[0185] Raw gene expression level and alternative splicing data may
in some cases be improved through the application of algorithms
designed to normalize and or improve the reliability of the data.
In some embodiments of the present invention the data analysis
requires a computer or other device, machine or apparatus for
application of the various algorithms described herein due to the
large number of individual data points that are processed. A
"machine learning algorithm" refers to a computational-based
prediction methodology, also known to persons skilled in the art as
a "classifier", employed for characterizing a gene expression
profile. The signals corresponding to certain expression levels,
which are obtained by, e.g., microarray-based hybridization assays,
are typically subjected to the algorithm in order to classify the
expression profile. Supervised learning generally involves
"training" a classifier to recognize the distinctions among classes
and then "testing" the accuracy of the classifier on an independent
test set. For new, unknown samples the classifier can be used to
predict the class in which the samples belong.
[0186] In some cases, the robust multi-array Average (RMA) method
may be used to normalize the raw data. The RMA method begins by
computing background-corrected intensities for each matched cell on
a number of microarrays. The background corrected values are
restricted to positive values as described by Irizarry et al.
Biostatistics 2003 Apr. 4 (2): 249-64. After background correction,
the base-2 logarithm of each background corrected matched-cell
intensity is then obtained. The back-ground corrected,
log-transformed, matched intensity on each microarray is then
normalized using the quantile normalization method in which for
each input array and each probe expression value, the array
percentile probe value is replaced with the average of all array
percentile points, this method is more completely described by
Bolstad et al. Bioinformatics 2003. Following quantile
normalization, the normalized data may then be fit to a linear
model to obtain an expression measure for each probe on each
microarray. Tukey's median polish algorithm (Tukey, J. W.,
Exploratory Data Analysis. 1977) may then be used to determine the
log-scale expression level for the normalized probe set data.
[0187] Data may further be filtered to remove data that may be
considered suspect. In some embodiments, data deriving from
microarray probes that have fewer than about 4, 5, 6, 7 or 8
guanosine+cytosine nucleotides may be considered to be unreliable
due to their aberrant hybridization propensity or secondary
structure issues. Similarly, data deriving from microarray probes
that have more than about 12, 13, 14, 15, 16, 17, 18, 19, 20, 21,
or 22 guanosine+cytosine nucleotides may be considered unreliable
due to their aberrant hybridization propensity or secondary
structure issues.
[0188] In some cases, unreliable probe sets may be selected for
exclusion from data analysis by ranking probe-set reliability
against a series of reference datasets. For example, RefSeq or
Ensembl (EMBL) are considered very high quality reference datasets.
Data from probe sets matching RefSeq or Ensembl sequences may in
some cases be specifically included in microarray analysis
experiments due to their expected high reliability. Similarly data
from probe-sets matching less reliable reference datasets may be
excluded from further analysis, or considered on a case by case
basis for inclusion. In some cases, the Ensembl high throughput
cDNA (HTC) and/or mRNA reference datasets may be used to determine
the probe-set reliability separately or together. In other cases,
probe-set reliability may be ranked. For example, probes and/or
probe-sets that match perfectly to all reference datasets such as
for example RefSeq, HTC, and mRNA, may be ranked as most reliable
(1). Furthermore, probes and/or probe-sets that match two out of
three reference datasets may be ranked as next most reliable (2),
probes and/or probe-sets that match one out of three reference
datasets may be ranked next (3) and probes and/or probe sets that
match no reference datasets may be ranked last (4). Probes and or
probe-sets may then be included or excluded from analysis based on
their ranking. For example, one may choose to include data from
category 1, 2, 3, and 4 probe-sets; category 1, 2, and 3
probe-sets; category 1 and 2 probe-sets; or category 1 probe-sets
for further analysis. In another example, probe-sets may be ranked
by the number of base pair mismatches to reference dataset entries.
It is understood that there are many methods understood in the art
for assessing the reliability of a given probe and/or probe-set for
molecular profiling and the methods of the present invention
encompass any of these methods and combinations thereof.
[0189] In some embodiments of the present invention, data from
probe-sets may be excluded from analysis if they are not expressed
or expressed at an undetectable level (not above background). A
probe-set is judged to be expressed above background if for any
group:
[0190] Integral from TO to Infinity of the standard normal
distribution<Significance (0.01)
Where:
[0191] T0=Sqr(GroupSize)(T-P)/Sqr(P var),
GroupSize=Number of CEL files in the group, T=Average of probe
scores in probe-set, P=Average of Background probes averages of GC
content, and
P var=Sum of Background probe variances/(Number of probes in
probe-set) 2,
[0192] This allows including probe-sets in which the average of
probe-sets in a group is greater than the average expression of
background probes of similar GC content as the probe-set probes as
the center of background for the probe-set and enables one to
derive the probe-set dispersion from the background probe-set
variance.
[0193] In some embodiments of the present invention, probe-sets
that exhibit no, or low variance may be excluded from further
analysis. Low-variance probe-sets are excluded from the analysis
via a Chi-Square test. A probe-set is considered to be low-variance
if its transformed variance is to the left of the 99 percent
confidence interval of the Chi-Squared distribution with (N-1)
degrees of freedom.
(N-1)*Probe-set Variance/(Gene Probe-set
Variance).about.Chi-Sq(N-1)
where N is the number of input CEL files, (N-1) is the degrees of
freedom for the Chi-Squared distribution, and the `probe-set
variance for the gene` is the average of probe-set variances across
the gene.
[0194] In some embodiments of the present invention, probe-sets for
a given gene or transcript cluster may be excluded from further
analysis if they contain less than a minimum number of probes that
pass through the previously described filter steps for GC content,
reliability, variance and the like. For example in some
embodiments, probe-sets for a given gene or transcript cluster may
be excluded from further analysis if they contain less than about
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or less than
about 20 probes.
[0195] Methods of data analysis of gene expression levels or of
alternative splicing may further include the use of a feature
selection algorithm as provided herein. In some embodiments of the
present invention, feature selection is provided by use of the
LIMMA software package (Smyth, G. K. (2005). Limma: linear models
for microarray data. In: Bioinformatics and Computational Biology
Solutions using R and Bioconductor, R. Gentleman, V. Carey, S.
Dudoit, R. Irizarry, W. Huber (eds.), Springer, New York, pages
397-420).
[0196] Methods of data analysis of gene expression levels and or of
alternative splicing may further include the use of a
pre-classifier algorithm. For example, an algorithm may use a
cell-specific molecular fingerprint to pre-classify the samples
according to their composition and then apply a
correction/normalization factor. This data/information may then be
fed in to a final classification algorithm which would incorporate
that information to aid in the final diagnosis.
[0197] Methods of data analysis of gene expression levels and or of
alternative splicing may further include the use of a classifier
algorithm as provided herein. In some embodiments of the present
invention a support vector machine (SVM) algorithm, a random forest
algorithm, or a combination thereof is provided for classification
of microarray data. In some embodiments, identified markers that
distinguish samples (e.g. benign vs. malignant, normal vs.
malignant) or distinguish subtypes (e.g. PTC vs. FVPTC) are
selected based on statistical significance. In some cases, the
statistical significance selection is performed after applying a
Benjamini Hochberg correction for false discovery rate (FDR).
[0198] In some cases, the classifier algorithm may be supplemented
with a meta-analysis approach such as that described by Fishel and
Kaufman et al. 2007 Bioinformatics 23(13): 1599-606. In some cases,
the classifier algorithm may be supplemented with a meta-analysis
approach such as a repeatability analysis. In some cases, the
repeatability analysis selects markers that appear in at least one
predictive expression product marker set.
[0199] In some cases, the results of feature selection and
classification may be ranked using a Bayesian post-analysis method.
For example, microarray data may be extracted, normalized, and
summarized using methods known in the art such as the methods
provided herein. The data may then be subjected to a feature
selection step such as any feature selection methods known in the
art such as the methods provided herein including but not limited
to the feature selection methods provided in LIMMA. The data may
then be subjected to a classification step such as any of the
classification methods known in the art such as the use of any of
the algorithms or methods provided herein including but not limited
to the use of SVM or random forest algorithms. The results of the
classifier algorithm may then be ranked by according to a posterior
probability function. For example, the posterior probability
function may be derived from examining known molecular profiling
results, such as published results, to derive prior probabilities
from type I and type II error rates of assigning a marker to a
category (e.g. benign, malignant, normal, ATC, PTC, MTC, FC, FN,
FA, FVPTC CN, HA, HC, LCT, NHP etc.). These error rates may be
calculated based on reported sample size for each study using an
estimated fold change value (e.g. 1.1, 1.2., 1.3, 1.4, 1.5, 1.6,
1.7, 1.8, 1.9, 2, 2.2, 2.4, 2.5, 3, 4, 5, 6, 7, 8, 9, 10 or more).
These prior probabilities may then be combined with a molecular
profiling dataset of the present invention to estimate the
posterior probability of differential gene expression. Finally, the
posterior probability estimates may be combined with a second
dataset of the present invention to formulate the final posterior
probabilities of differential expression. Additional methods for
deriving and applying posterior probabilities to the analysis of
microarray data are known in the art and have been described for
example in Smyth, G. K. 2004 Stat. Appl. Genet. Mol. Biol. 3:
Article 3. In some cases, the posterior probabilities may be used
to rank the markers provided by the classifier algorithm. In some
cases, markers may be ranked according to their posterior
probabilities and those that pass a chosen threshold may be chosen
as markers whose differential expression is indicative of or
diagnostic for samples that are for example benign, malignant,
normal, ATC, PTC, MTC, FC, FN, FA, FVPTC CN, HA, HC, LCT, or NHP.
Illustrative threshold values include prior probabilities of 0.7,
0.75, 0.8, 0.85, 0.9, 0.925, 0.95, 0.975, 0.98, 0.985, 0.99, 0.995
or higher.
[0200] A statistical evaluation of the results of the molecular
profiling may provide a quantitative value or values indicative of
one or more of the following: the likelihood of diagnostic
accuracy, the likelihood of cancer, disease or condition, the
likelihood of a particular cancer, disease or condition, the
likelihood of the success of a particular therapeutic intervention.
Thus a physician, who is not likely to be trained in genetics or
molecular biology, need not understand the raw data. Rather, the
data is presented directly to the physician in its most useful form
to guide patient care. The results of the molecular profiling can
be statistically evaluated using a number of methods known to the
art including, but not limited to: the students T test, the two
sided T test, pearson rank sum analysis, hidden markov model
analysis, analysis of q-q plots, principal component analysis, one
way ANOVA, two way ANOVA, LIMMA and the like.
[0201] In some embodiments of the present invention, the use of
molecular profiling alone or in combination with cytological
analysis may provide a diagnosis that is between about 85% accurate
and about 99% or about 100% accurate. In some cases, the molecular
profiling business may through the use of molecular profiling
and/or cytology provide a diagnosis of malignant, benign, or normal
that is about 85%, 86%, 87%, 88%, 90%, 91%, 92%, 93%, 94%, 95%,
96%, 97%, 97.5%, 98%, 98.5%, 99%, 99.5%, 99.75%, 99.8%, 99.85%, or
99.9% accurate.
[0202] In some cases, accuracy may be determined by tracking the
subject over time to determine the accuracy of the original
diagnosis. In other cases, accuracy may be established in a
deterministic manner or using statistical methods. For example,
receiver operator characteristic (ROC) analysis may be used to
determine the optimal assay parameters to achieve a specific level
of accuracy, specificity, positive predictive value, negative
predictive value, and/or false discovery rate. Methods for using
ROC analysis in cancer diagnosis are known in the art and have been
described for example in US Patent Application No. 2006/019615
herein incorporated by reference in its entirety.
[0203] In some embodiments of the present invention, gene
expression products and compositions of nucleotides encoding for
such products which are determined to exhibit the greatest
difference in expression level or the greatest difference in
alternative splicing between benign and normal, benign and
malignant, or malignant and normal may be chosen for use as
molecular profiling reagents of the present invention. Such gene
expression products may be particularly useful by providing a wider
dynamic range, greater signal to noise, improved diagnostic power,
lower likelihood of false positives or false negative, or a greater
statistical confidence level than other methods known or used in
the art.
[0204] In other embodiments of the present invention, the use of
molecular profiling alone or in combination with cytological
analysis may reduce the number of samples scored as non-diagnostic
by about 100%, 99%, 95%, 90%, 80%, 75%, 70%, 65%, or about 60% when
compared to the use of standard cytological techniques known to the
art. In some cases, the methods of the present invention may reduce
the number of samples scored as intermediate or suspicious by about
100%, 99%, 98%, 97%, 95%, 90%, 85%, 80%, 75%, 70%, 65%, or about
60%, when compared to the standard cytological methods used in the
art.
[0205] In some cases the results of the molecular profiling assays,
are entered into a database for access by representatives or agents
of the molecular profiling business, the individual, a medical
provider, or insurance provider. In some cases assay results
include interpretation or diagnosis by a representative, agent or
consultant of the business, such as a medical professional. In
other cases, a computer or algorithmic analysis of the data is
provided automatically. In some cases the molecular profiling
business may bill the individual, insurance provider, medical
provider, researcher, or government entity for one or more of the
following: molecular profiling assays performed, consulting
services, data analysis, reporting of results, or database
access.
[0206] In some embodiments of the present invention, the results of
the molecular profiling are presented as a report on a computer
screen or as a paper record. In some cases, the report may include,
but is not limited to, such information as one or more of the
following: the number of genes differentially expressed, the
suitability of the original sample, the number of genes showing
differential alternative splicing, a diagnosis, a statistical
confidence for the diagnosis, the likelihood of cancer or
malignancy, and indicated therapies.
[0207] (iv) Categorization of Samples Based on Molecular Profiling
Results
[0208] The results of the molecular profiling may be classified
into one of the following: benign (free of a cancer, disease, or
condition), malignant (positive diagnosis for a cancer, disease, or
condition), or non diagnostic (providing inadequate information
concerning the presence or absence of a cancer, disease, or
condition). In some cases, a diagnostic result may further classify
the type of cancer, disease or condition. In other cases, a
diagnostic result may indicate a certain molecular pathway involved
in the cancer disease or condition, or a certain grade or stage of
a particular cancer disease or condition. In still other cases a
diagnostic result may inform an appropriate therapeutic
intervention, such as a specific drug regimen like a kinase
inhibitor such as Gleevec or any drug known to the art, or a
surgical intervention like a thyroidectomy or a
hemithyroidectomy.
[0209] In some embodiments of the present invention, results are
classified using a trained algorithm. Trained algorithms of the
present invention include algorithms that have been developed using
a reference set of known malignant, benign, and normal samples
including but not limited to the samples listed in FIG. 1.
Algorithms suitable for categorization of samples include but are
not limited to k-nearest neighbor algorithms, concept vector
algorithms, naive bayesian algorithms, neural network algorithms,
hidden markov model algorithms, genetic algorithms, and mutual
information feature selection algorithms or any combination
thereof. In some cases, trained algorithms of the present invention
may incorporate data other than gene expression or alternative
splicing data such as but not limited to DNA polymorphism data,
sequencing data, scoring or diagnosis by cytologists or
pathologists of the present invention, information provided by the
pre-classifier algorithm of the present invention, or information
about the medical history of the subject of the present
invention.
[0210] (v) Monitoring of Subjects or Therapeutic Interventions Via
Molecular Profiling
[0211] In some embodiments, a subject may be monitored using
methods and compositions of the present invention. For example, a
subject may be diagnosed with cancer or a genetic disorder. This
initial diagnosis may or may not involve the use of molecular
profiling. The subject may be prescribed a therapeutic intervention
such as a thyroidectomy for a subject suspected of having thyroid
cancer. The results of the therapeutic intervention may be
monitored on an ongoing basis by molecular profiling to detect the
efficacy of the therapeutic intervention. In another example, a
subject may be diagnosed with a benign tumor or a precancerous
lesion or nodule, and the tumor, nodule, or lesion may be monitored
on an ongoing basis by molecular profiling to detect any changes in
the state of the tumor or lesion.
[0212] Molecular profiling may also be used to ascertain the
potential efficacy of a specific therapeutic intervention prior to
administering to a subject. For example, a subject may be diagnosed
with cancer. Molecular profiling may indicate the upregulation of a
gene expression product known to be involved in cancer malignancy,
such as for example the RAS oncogene. A tumor sample may be
obtained and cultured in vitro using methods known to the art. The
application of various inhibitors of the aberrantly activated or
dysregulated pathway, or drugs known to inhibit the activity of the
pathway may then be tested against the tumor cell line for growth
inhibition. Molecular profiling may also be used to monitor the
effect of these inhibitors on for example down-stream targets of
the implicated pathway.
[0213] (vi) Molecular Profiling as a Research Tool
[0214] In some embodiments, molecular profiling may be used as a
research tool to identify new markers for diagnosis of suspected
tumors; to monitor the effect of drugs or candidate drugs on
biological samples such as tumor cells, cell lines, tissues, or
organisms; or to uncover new pathways for oncogenesis and/or tumor
suppression.
[0215] (vii) Biomarker Groupings Based on Molecular Profiling
[0216] Thyroid genes are described according to the groups 1)
Benign vs. Malignant, 2) alternative gene splicing, 3) KEGG
Pathways, 4) Normal Thyroid, 5) Thyroid pathology subtype, 6) Gene
Ontology, and 7) Biomarkers of metastasis to the thyroid from
non-thyroid organs. Methods and compositions of the invention can
have genes selected from one or more of the groups listed above
and/or 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45,
50 or more subgroups from any of the groups listed above (e.g. one
or more different KEGG pathway) and can have from 1, 2, 3, 4, 5, 6,
7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more gene expression
products from each group, in any combination. In some embodiments,
the set of genes combined give a specificity or sensitivity of
greater than 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%,
93%, 94%, 95%, 96%, 97%, 98%, 99%, or 99.5%, or a positive
predictive value or negative predictive value of at least 95%,
95.5%, 96%, 96.5%, 97%, 97.5%, 98%, 98.5%, 99%, 99.5% or more.
[0217] In some embodiments, the extracellular matrix, adherens,
focal adhesion, and tight junction genes are used as biomarkers of
thyroid cancer. In some embodiments, the signaling pathway is
selected from one of the following three pathways: adherens
pathway, focal adhesion pathway, and tight junction pathway. In
some embodiments, at least one gene is selected from one of the 3
pathways. In some embodiments, at least one gene is selected from
each one of the three pathways. In some embodiments, at least one
gene is selected from two of the three pathways. In some
embodiments, at least one gene that is involved in all three
pathways is selected. In one example, a set of genes that is
involved in adherens pathway, focal adhesion pathway, and tight
junction pathway is selected as the markers for diagnosis of a
cancer such as thyroid cancer.
[0218] The follicular cells that line thyroid follicles are highly
polarized and organized in structure, requiring distinct roles of
their luminal and apical cell membranes. In some embodiments,
cytoskeleton, plasma membrane, and extracellular space genes are
used as biomarkers of thyroid cancer. In some embodiments, genes
that overlap all four pathways, i.e. ECM, focal adhesion, adherens,
and tight junction pathways, are used as biomarkers of thyroid
cancer. In one example, the present invention provides the Benign
vs. malignant group (n=948) as a thyroid classification gene list.
This list has been grouped according to alternative splicing, KEGG
pathways, and gene ontology. KEGG pathways are further described in
Table 1.
[0219] In some embodiments, the present invention provides a method
of diagnosing cancer comprising gene expression products from one
or more signaling pathways that include but are not limited to the
following: acute myeloid leukemia signaling, somatostatin receptor
2 signaling, cAMP-mediated signaling, cell cycle and DNA damage
checkpoint signaling, G-protein coupled receptor signaling,
integrin signaling, melanoma cell signaling, relaxin signaling, and
thyroid cancer signaling. Methods and compositions of the invention
can have genes selected from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20,
25, 30, 35, 40, 45, 50 or more signaling pathways and can have from
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 or
more gene expression products from each signaling pathway, in any
combination. In some embodiments, the set of genes combined give a
specificity or sensitivity of greater than 70%, 75%, 80%, 85%, 86%,
87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or
99.5%, or a positive predictive value or negative predictive value
of at least 95%, 95.5%, 96%, 96.5%, 97%, 97.5%, 98%, 98.5%, 99%,
99.5% or more.
[0220] In some embodiments, the present invention provides a method
of diagnosing cancer comprising gene expression products from one
or more ontology groups that include but are not limited to the
following: cell aging, cell cortex, cell cycle, cell
death/apoptosis, cell differentiation, cell division, cell
junction, cell migration, cell morphogenesis, cell motion, cell
projection, cell proliferation, cell recognition, cell soma, cell
surface, cell surface linked receptor signal transduction, cell
adhesion, transcription, immune response, or inflammation. Methods
and compositions of the invention can have genes selected from 1,
2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more
ontology groups and can have from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
15, 20, 25, 30, 35, 40, 45, 50 or more gene expression products
from each ontology group, in any combination. In some embodiments,
the set of genes combined give a specificity or sensitivity of
greater than 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%,
93%, 94%, 95%, 96%, 97%, 98%, 99%, or 99.5%, or a positive
predictive value or negative predictive value of at least 95%,
95.5%, 96%, 96.5%, 97%, 97.5%, 98%, 98.5%, 99%, 99.5% or more.
TABLE-US-00001 TABLE 1 Genes involved in the KEGG Pathways Genes %
in Top 948 B in Top 948 B Total Genes in KEGG Pathway vs. M list
vs. M list Pathway ECM 23 18 84 p53 14 10 69 PPAR 14 10 69 Thyroid
Cancer 14 4 29 Focal Adhesion 13 26 201 Adherens 12 9 77 Tight
Junction 11 14 134 Pathways in Cancer 10 33 332 Overview Jak/STAT
10 14 155 Cell Cycle 7 9 129 TGFbeta 7 6 87 Wnt 7 10 151 ErbB 6 5
87 Apoptosis 6 5 88 MAPK 5 14 269 Autoimmune Thyroid 4 2 53 mTOR 2
1 53 VEGF 1 1 76
[0221] Top Biomarkers of benign vs. malignant thyroid, n=948, are
listed below in List 1:
List 1
[0222] TCID-2406391, TCID-3153400, TCID-3749600, ABCC3, ABCD2,
ABTB2, ACBD7, ACSL1, ACTA2, ADAMTS5, ADAMTS9, ADK, ADORA1, AEBP1,
AFAP1, AGR2, AHNAK2, AHR, AIDA, AMU, AK1, AKR1C3, ALAS2, ALDH1A3,
ALDH1B1, ALDH6A1, ALOX5, AMIGO2, AMOT, ANGPTL1, ANK2, ANKS6, ANO5,
ANXA1, ANXA2, ANXA2P1, ANXA3, ANXA6, AOAH, AP3S1, APOBEC3F,
APOBEC3G, APOL1, APOO, AQP4, AQP9, ARHGAP19, ARHGAP24, ARL13B,
ARL4A, ARMCX3, ARMCX6, ARNTL, ARSG, ASAP2, ATIC, ATM, ATP13A4,
ATP6VOD2, ATP8A1, AUTS2, AVPR1A, B3GNT3, BAG3, BCL2, BCL2A1, BCL9,
BHLHE40, BHLHE41, MRCS, BLNK, BMP1, BMP8A, BTBD11, BTG3, C10orf131,
C10orf72, C11orf72, C11orf74, C11orf80, C12orf35, C12orf49,
C14orf45, C16orf45, C17orf87, C19orf33, C1orf115, C1orf116, C2,
C22orf9, C2orf40, C3, C4A, C4B, C4orf34, C4orf7, C5orf28, C6orf168,
C6orf174, C7orf62, C8orf16, C8orf39, C8orf4, C8orf79, C9orf68,
CA11, CADM1, CALCA, CAMK2N1, CAMK4, CAND1, CARD16, CARD17, CARD8,
CASC5, CASP1, CAV1, CAV2, CCDC109B, CCDC121, CCDC146, CCDC148,
CCDC152, CCDC80, CCL13, CCL19, CCND1, CCND2, CD151, CD180, CD2,
CD200, CD36, CD3D, CD48, CD52, CD69, CD79A, CD96, CDCP1, CDH11,
CDH3, CDH6, CDK2, CDKL2, CDO1, CDON, CDR1, CEP110, CEP55, CERKL,
CFB, CFH, CFHR1, CFI, CHAF1B, CHD4, CHGB, CHI3L1, CITED1, CKB,
CKS2, CLC, CLDN1, CLDN10, CLDN16, CLDN4, CLDN7, CLEC2B, CLEC4E,
CLIP3, CLU, CMAH, CNN2, CNN3, COL12A1, COL1A1, COPZ2, CP, CPE,
CPNE3, CR2, CRABP1, CRABP2, CSF3R, CSGALNACT1, CST6, CTNNAL1,
CTNNB1, CTSC, CTSH, CTTN, CWH43, CXCL1, CXCL11, CXCL13, CXCL14,
CXCL17, CXCL2, CXCL3, CXCL9, CXorf18, CXorf27, CYP1B1, CYP24A1,
CYP27A1, CYP4B1, CYSLTR1, CYSLTR2, CYTH1, DAPK2, DCAF17, DCBLD2,
DCUN1D3, DDAH1, DDB2, DDX52, DENND4A, DGKH, DGKI, DHRS1, DHRS3,
DI01, DIRAS3, DLC1, DLG2, DLG4, DLGAP5, DNAJB14, DNASE1L3, DOCK8,
DOCK9, DOK4, DPH3B, DPP4, DPYD, DPYSL3, DSG2, DSP, DST, DUOX1,
DUOX2, DUOXA1, DUOXA2, DUSP4, DUSP5, DUSP6, DYNC1I2, DYNLT1, DZIP1,
ECE1, EDNRB, EFEMP1, EGF, EGFR, EHBP1, EHD2, EHF, EIF2B2, EIF4H,
ELK3, ELMO1, EMP2, EMR3, ENAH, ENDOD1, ENTPD1, EPB41, EPDR1, EPHA4,
EPHX4, EPR1, EPS8, ERBB2, ERBB3, ERI2, ERO1LB, ERP27, ESRRG, ETNK2,
ETS1, ETV1, ETV4, ETV5, F2RL2, F8, FAAH2, FABP4, FAM111A, FAM111B,
FAM164A, FAM176A, FAM20A, FAM55C, FAM82B, FAM84B, FAT4, FBLN5,
FBXO2, FBXO21, FCN1, FCN2, FGF2, FGFR10P2, FIB1N, FLJ20184,
F1126056, FL132810, FLJ42258, FLRT3, FN1, FPR1, FPR2, FREM2, FRMD3,
FXYD6, FYB, FZD4, FZD6, FZD7, GOS2, GABBR2, GABRB2, GADD45A, GALE,
GALNT12, GALNT3, GALNT7, GBE1, GBP1, GBP3, GBP5, GGCT, GIMAP2,
GIMAP5, GIMAP7, GJA4, GLA, GLDC, GLDN, GLIS3, GNG12, GOLT1A, GPAM,
GPR110, GPR125, GPR155, GPR174, GPR98, GPRC5B, GRAMD3, GSN, GTF3A,
GULP1, GYPB, GYPC, GYPE, GZMA, GZMK, HEMGN, HEY2, HIGD1A, HIPK2,
HIST1H1A, HIST1H3B, HIST1H4L, HK1, HLA-DPB1, HLA-DQB2, HLF, HMGA2,
HMMR, HNRNPM, HPN, HPS3, HRASLS, HSD17B6, HSPH1, ICAM1, ID3, IFI16,
IFITM1, IFNAR2, IGF2BP2, IGFBP5, IGFBP6, IGFBP7, IGJ, IGK, IGKC,
IGKV1-5, IGKV3-15, IGKV3-20, IGKV3D-11, IGKV3D-15, IGSF1, IKZF2,
IKZF3, IKZF4, IL1RAP, IL1RL1, IL2RA, IL7R, IL8, IL8RA, IL8RB,
IL8RBP, IMPDH2, INPP5F, IPCEF1, IQGAP2, ISYNA1, ITGA2, ITGA3,
ITGA4, ITGA9, ITGB1, ITGB4, ITGB6, ITGB8, ITM2A, ITPR1, IYD, JAK2,
JUB, KAL1, KATNAL2, KBTBD8, KCNA3, KCNABI, KCNK5, KCNQ3, KCTD14,
KDELC1, KDELR3, KHDRBS2, KIAA0284, KIAA0408, KIAAl217, KIAA1305,
KIF11, KIT, KLF8, KLHDC8A, KLHL6, KLK10, KLK7, KLRB1, KLRC4, KLRG1,
KLRK1, KRT18, KRT19, KYNU, LAMB1, LAMB3, LAMC1, LAMC2, LCA5, LCMT1,
LCN2, LCP1, LDOC1, LEMD1, LGALS2, LGALS3, LIFR, LILRA1, LILRB1,
LIMA1, LINGO2, LIPH, LMO3, LMO4, LOC100124692, LOC100127974,
LOC100129112, LOC100129115, LOC100129171, LOC100129961,
LOC100130100, LOC100130248, LOC100131102, LOC100131490,
LOC100131869, LOC100131938, LOC100131993, LOC100132338,
LOC100132764, LOC26080, LOC283508, LOC284861, LOC439911, LOC440434,
LOC440871, L00554202, LOC643454, LOC646358, LOC648149, LOC650405,
LOC652493, LOC652694, LOC653264, LOC653354, LOC653498, LOC728212,
LOC729461, LOC730031, LONRF2, LOX, LPAR1, LPAR5, LPCAT2, LPL,
LRP1B, LRP2, LRRC69, LRRN1, LRRN3, LTBP2, LTBP3, LUM, LYPLA1,
LYRM1, LYZ, MACC1, MAFG, MAGOH2, MAMLD1, MAP2, MAPK4, MAPK6, MATN2,
MBOAT2, MCM4, MCMI, MDK, ME1, MED13, MED13L, MELK, MET, METTL7B,
MEX3C, MFGE8, MGAM, MGAT1, MGAT4C, MGC2889, MGST1, MIS12, MKI67,
MLLT3, MLLT4, MMP16, MNDA, MORC4, MPPED2, MPZL2, MRC2, MRPL14,
MT1F, MT1G, MT1H, MT1M, MT1P2, MT1P3, MTHFD1L, MTIF3, MUC1, MUC15,
MVP, MXRA5, MYEF2, MYH10, MY01B, MY01D, MY05A, MY06, NAB2, NAE1,
NAG20, NAV2, NCAM1, NCKAP1, ND1, NDC80, NDFIP2, NEB, NEDD4L, NELL2,
NEXN, NFATC3, NFE2, NFIB, NFKBIZ, NIPAL3, NIPSNAP3A, NIPSNAP3B,
NOD1, NPAS3, NPAT, NPC2, NPEPPS, NPL, NPY1R, NRCAM, NRIP1, NRP2,
NT5E, NTAN1, NUCB2, NUDT6, NUPR1, NUSAP1, OCIAD2, OCR1, ODZ1,
ORAOV1, OSBPL1A, OSGEP, OSMR, P2RY13, P4HA2, PAM, PAPSS2, PARD6B,
PARP14, PARP4, PARVA, PBX1, PCDH1, PCMTD1, PCNXL2, PDE5A, PDE9A,
PDGFRL, PDK4, PDLIM1, PDLIM4, PDZRN4, PEG10, PERP, PGCP, PHEX,
PHF16, PHLDB2, PHYHIP, PIAS3, PIGN, PKHD1L1, PKP2, PKP4, PLA2G16,
PLA2G7, PLA2R1, PLAG1, PLAU, PLCD3, PLCL1, PLEK, PLEKHA4, PLEKHA5,
PLEKHF2, PLK2, PLP2, PLS3, PLSCR4, PLXNC1, PMEPA1, POLR2J4, PON2,
POR, POU2F3, PPAP2C, PPARGC1A, PPBP, PPL, PPP1R14C, PRCP, PRICKLE1,
PRINS, PRMT6, PROK2, PROS1, PRR15, PRRG1, PRSS23, PSAT1, PSD3,
PTK7, PTPN14, PTPN22, PTPRC, PTPRE, PTPRF, PTPRG, PTPRK, PTPRU,
PTRF, PXDNL, PYGL, PYHIN1, QTRT1, RAB25, RAB27A, RAB32, RAB34,
RAD23B, RAG2, RAI2, RAPGEF5, RARG, RASA1, RASD2, RBBP7, RBBP8,
RBMS2, RCBTB2, RCE1, RDH5, RG9MTD2, RGS13, RGS18, RGS2, RHOBTB3,
RHOH, RHOU, RICH2, RIMS2, RNASE1, RNASET2; RND3, ROS1, RPL39L,
RPL9P11, RPRD1A, RPS6KA6, RRAS, RRAS2, RRBP1, RRM2, RUNX1, RUNX2,
RXRG, RYR2, S100A12, S100A14, S100A16, S100A8, S100A9, SALL1, SAV1,
SC4MOL, SCARA3, SCARNA11, SCEL, SCG3, SCG5, SCNN1A, SCP2, SCRN1,
SDC4, SDK1, SEH1L, SEL1L3, SELL, SEMA3C, SEMA3D, SEMA4C, SEPP1,
SEPT11, SERGEF, SERINC2, SERPINA1, SERPINA2, SERPINE2, SERPING1,
SFN, SFTPB, SGCB, SGCE, SGEF, SGMS2, SGPP2, SH2D4A, SH3BGR,
SH3PXD2A, SIPA1L2, SIRPA, SIRPB1, SLA, SLC12A2, SLC16A4, SLC16A6,
SLC17A5, SLC24A5, SLC25A33, SLC26A4, SLC26A7, SLC27A2, SLC27A6,
SLC34A2, SLC35D2, SLC35F2, SLC39A6, SLC4A4, SLC5A8, SLC7A11,
SLC7A2, SLIT1, SLIT2, SLPI, SMAD9, SMOC2, SMURF2, SNCA, SNX1,
SNX22, SNX7, SOAT1, SORBS2, SP140, SP140L, SPATS2, SPATS2L, SPC25,
SPINT1, SPOCK1, SPP1, SPRED2, SPRY1, SPRY2, SQLE, SRL, SSPN, ST20,
ST3GAL5, STAT4, STEAP2, STK17B, STK32A, STXBP6, SULF1, SYNE1,
SYT14, SYTL5, TACSTD2, TASP1, TBC1D3F, TC2N, TCERG1L, TCF7L2,
TCFL5, TDRKH, TEAD1, TFCP2L1, TFF3, TFPI, TGFA, TGFB2, TGFBR1,
THSD4, TIAM2, TIMP1, TIMP3, TIPARP, TJP1, TJP2, TLCD1, TLE4, TLR10,
TLR8, TM4SF1, TM4SF4, TM7SF4, TMEM100, TMEM117, TMEM133, TMEM156,
TMEM163, TMEM171, TMEM215, TMEM220, TMEM90A, TMEM98, TMPRSS4,
TMSB10, TMSB15A, TMSB15B, TNC, TNFAIP8, TNFRSF11B, TNFRSF12A,
TNFRSF17, TNFSF10, TNFSF15, TOMM34, TOX, TPD52L1, TPO, TPX2,
TRIP10, TRPC5, TRPC6, TSC22D1, TSHZ2, TSPAN13, TSPAN6, TSPAN8,
TSSC1, TTC39A, TUBB1, TUBB6, TULP3, TUSC3, TXNL1, TXNRD1, TYMS,
UCHL5, VAMPI, VNN1, VNN2, VNN3, WDR40A, WDR54, WDR72, WIPI1, WNT5A,
XKRX, XPR1, YIF1B, YIPF1, YTHDC2, ZBTB33, ZCCHC12, ZCCHC16, ZEB2,
ZFP36L1, ZFPM2, ZMAT3, ZMAT4, ZNF143, ZNF208, ZNF487, ZNF643,
ZNF804B, ZYG11A.
[0223] Alternative spliced genes, n=283, are listed below in List
2:
List 2
[0224] ABCC3, ADAMTS5, ADAMTS9, AIDA, AK1, AKR1C3, ALDH1A3,
ALDH6A1, AMIG02, AMOT, ANGPTL1, ANKS6, ANO5, ANXA1, ANXA2, ANXA2P1,
ANXA3, AQP4, ARHGAP24, ARL4A, ARMCX3, ARMCX6, ARSG, ATIC, ATP13A4,
ATP8A1, AUTS2, BAG3, BCL2, BCL9, BHLHE41, C10orf131, C11orf74,
C14orf45, C16orf45, C19orf33, C2orf40, C3, C5orf28, C8orf79, CA11,
CALCA, CAV1, CCND1, CCND2, CD36, CD36, CDH3, CDH6, CDON, CFH,
CFHR1, CHD4, CITED1, CLDN16, CLU, COPZ2, CP, CRABP1, CSGALNACT1,
CTSC, CTSH, CTTN, CWH43, CYSLTR2, DCBLD2, DCUN1D3, DDB2, DGKH,
DGK1, DIO1, DLG2, DOCK9, DPH3B, DPP4, DSP, DST, DUSP6, EFEMP1,
EIF2B2, ELMO1, EMP2, ENAH, ENTPD1, EPHX4, ERBB3, ERI2, ERO1LB,
ETNK2, ETV1, ETV5, F8, FABP4, FAM111B, FAM20A, FAM55C, FAT4, FBLN5,
FGFR10P2, FLJ42258, FLRT3, FN1, FREM2, FXYD6, GABBR2, GABRB2,
GALNT7, GBE1, GBP1, GBP3, GGCT, GIMAP7, GPAM, GPR125, GPR155,
GRAMD3, GSN, HLF, HMGA2, HSPH1, IMPDH2, IQGAP2, ITGA2, ITGA3,
ITGA9, ITGB6, ITGB8, ITM2A, ITPR1, IYD, KATNAL2, KCNA3, KCNQ3,
KDELC1, KHDRBS2, KIAA0284, KIAA1217, KIT, KLF8, KLK10, KRT19,
LAMBS, LAMC2, LEMD1, LIFR, LINGO2, LMO3, LOC100127974,
LOC100129112, LOC100131490, LOC100131869, LOC283508, LOC648149,
LOC653354, LONRF2, LPCAT2, LPL, LRP1B, LRP2, LRRC69, LRRN1, LRRN3,
LYRM1, MACC1, MAFG, MAP2, MAPK4, MAPK6, MATN2, MED13, MET, METTL7B,
MFGE8, MLLT3, MPPED2, MPZL2, MRPL14, MT1F, MT1G, MT1H, MT1P2,
MTHFD1L, MUC1, MVP, MYEF2, MYH10, MYO1D, NAG20, NAV2, NEB, NEDD4L,
NELL2, NFATC3, NFKBIZ, NPC2, NRCAM, NUCB2, ORAOV1, P4HA2, PAM,
PAPSS2, PARVA, PDLIM4, PEG10, PGCP, PIGN, PKHD1L1, PLA2G16, PLA2G7,
PLA2R1, PLAU, PLEKHA4, PLP2, PLSCR4, PLXNC1, PMEPA1, PON2,
PPARGC1A, PRINS, PROS1, PSD3, PTPRK, PYH1N1, QTRT1, RAB27A, RAB34,
RAD23B, RASA1, RHOBTB3, RNASET2, RPS6KA6, RUNX1, SCARNA11, SCG5,
SDC4, SERPINA1, SERPINA2, SGEF, SH2D4A, SLA, SLC12A2, SLC24A5,
SLC26A4, SLC26A7, SLC27A2, SLC27A6, SLC35F2, SLC4A4, SLC5A8,
SLC7A2, SOAT1, SPATS2, SPATS2L, SPINT1, SPP1, SSPN, STK32A, SULF1,
SYNE1, TCFL5, TFPI, TGFBR1, TIPARP, TJP1, TLE4, TM7SF4, TMEM171,
TMEM90A, TNFAIP8, TNFRSF11B, TOMM34, TPD52L1, TPO, TSC22D1, TUSC3,
TYMS, WDR54, WDR72, WIPI1, XPR1, YIF1B, ZFPM2, ZMAT4.
[0225] Genes involved in the KEGG pathways are listed below in
Table 6: there are 18 pathways with a total of n=109 unique
genes.
TABLE-US-00002 TABLE 6 Number of Signaling Pathway Genes Genes ECM
Pathway 19 CD36, COL1A1, FN1, HMMR, ITGA2, ITGA3, ITGA4, ITGA9,
ITGB1, ITGB4, ITGB6, ITGB8, LAMB1, LAMB3, LAMC1, LAMC2, SDC4, SPP1,
TNC p53 Pathway 10 ATM, CCND1, CCND2, CDK2, DDB2, GADD45A, PERP,
RRM2, SFN, ZMAT3 PPAR Pathway 10 ACSL1, CD36, CYP27A1, FABP4, LPL,
ME1, RXRG, SCP2, SLC27A2, SLC27A6 Thyroid Cancer 4 CCND1, CTNNB1,
RXRG, Pathway TCF7L2 Focal Adhesion 26 BCL2, CAV1, CAV2, CCND1,
Pathway CCND2, COL1A1, CTNNB1, EGF, EGFR, ERBB2, FN1, ITGA2, ITGA3,
ITGA4, ITGA9, ITGB1, ITGB6, ITGB8, LAMB1, LAMB3, LAMC1, LAMC2, MET,
PARVA, SPP1, TNC Adherens Pathway 9 CTNNB1, EGFR, ERBB2, MET,
MLLT4, PTPRF, TCF7L2, TGFBR1, TJP1 Tight Junctions 15 CLDN1,
CLDN10, CLDN16, Pathway CLDN4, CLDN7, CTNNB1, CTTN, EPB41, MLLT4,
MYH10, PARD6B, RRAS, RRAS2, TJP1, TJP2 Pathways in Cancer 34 BCL2,
BIRC5, CCND1, CDK2, Overview CSF3R, CTNNB1, DAPK2, EGF, EGFR,
ERBB2, ETS1, FGF2, FN1, FZD4, FZD6, FZD7, IL8, ITGA2, ITGA3, ITGB1,
KIT, LAMB1, LAMB3, LAMC1, LAMC2, MET, PIAS3, RUNX1, RXRG, TCF7L2,
TGFA, TGFB2, TGFBR1, WNT5A Jak/STAT Pathway 16 CCND1, CCND2, CSF3R,
IFNAR2, IL2RA, IL7R, ITGB4, JAK2, LIFR, OSMR, PIAS3, SPRED2, SPRY1,
SPRY2, STAT4, TPO Cell Cycle Pathway 9 ATM, CCND1, CCND2, CDK2,
GADD45A, MCM4, MCM7, SFN, TGFB2 TGFbeta Pathway 6 BMP8A, ID3,
SMAD9, SMURF2, TGFB2, TGFBR1 Wnt Pathway 10 CCND1, CCND2, CTNNB1,
FZD4, FZD6, FZD7, NFATC3, PRICKLE1, TCF7L2, WNT5A Erb Pathway 5
EGF, EGFR, ERBB2, ERBB3, TGFA Apoptosis Pathway 5 ATM, BCL2,
ENDOD1, IL1RAP, TNFSF10 MAPK Pathway 14 DUSP4, DUSP5, DUSP6, EGF,
EGFR, FGF2, GADD45A, GNG12, RASA1, RPS6KA6, RRAS, RRAS2, TGFB2,
TGFBR1 Autoimmune 2 HLA-DPB1, TPO Thyroid pathway mTOR Pathway 1
RPS6KA6 VEGF Pathway 1 NFATC3
[0226] Top genes separating benign and malignant thyroid (combined)
from normal thyroid, n=55, are listed below in List 3:
List 3
[0227] ANGPTL1, ANXA3, C10orf131, C2orf40, C7orf62, CAV1, CCDC80,
CDR1, CFH, CFHR1, CLDN16, CP, CRABP1, EFEMP1, ENTPD1, FABP4, FBLN5,
FN1, GBP1, GBP3, GULP1, HSD17B6, IPCEF1, KIT, LRP1B, LRRC69, LUM,
MAPK6, MATN2, MPPED2, MT1F, MT1G, MT1H, MT1M, MT1P2, MT1P3, MYEF2,
NRCAM, ODZ1, PAPSS2, PKHD1L1, PLA2R1, RYR2, SEMA3D, SLC24A5,
SLC26A4, SLC26A7, SLIT2, TFPI, TMEM171, TPO, TSPAN8, YTHDC2, ZFPM2,
ZNF804B.
[0228] Thyroid surgical pathology subtypes, n=873, are listed
below:
[0229] (i) List 4: FA subtype, n=243:
[0230] TM-3124344, AHR, ALOX5, ANGPTL1, ANXA2, ANXA2P1, APOL1,
AVPR1A, BMP8A, BTBD11, C2, C3, C8orf39, CCDC109B, CD36, CDON, CFB,
CHGB, CHI3L1, CKB, CLDN1, CP, CRABP1, CTSC, CTSH, CXCL1, CXCL2,
CXCL3, CXorf27, CYP1B1, DLG2, DNASE1L3, DPP4, DUOX1, DUOX2, DYNLT1,
EIF4H, F8, FABP4, FAM20A, FAM55C, FBLN5, FLJ26056, FXYD6, GOS2,
GALNT7, GLIS3, GPAM, HIGD1A, HKI, HLF, HSD17B6, ICAM1, IGFBP7,
ILIRAP, IPCEF1, IYD, KATNAL2, KCNAB1, KHDRBS2, KLF8, KLHDC8A,
LAMB1, LGALS3, LOC100131869, LOC26080, LOC284861, LOC439911,
LOC653264, LOC728212, LOC729461, LPCAT2, LRRC69, MAGOH2, MAPK4,
MAPK6, MELK, MPPED2, MT1G, NEB, NFKBIZ, NRIP1, PARP14, PKHD1L1,
PLA2G7, PLP2, PLXNC1, POR, PRMT6, PROS1, PSMB2, PTPRE, PYGL,
RNASE1, RNASET2, RPL9P11, RRAS2, RRBP1, RUNX1, RUNX2, RYR2, SCP2,
SEL1L3, SERGEF, SGPP2, SH3BGR, SLC25A33, SLC26A4, SLC26A7, SLC27A6,
SLC4A4, SLPI, SORBS2, SQLE, STK32A, SYTL5, TFCP2L1, TIAM2, TIMP3,
TMEM220, TMSB10, TRPC6, TSHZ2, TSSC1, VAMP1, ZNF487, ABCC3,
C11orf72, C8orf79, CLDN16, CLU, CST6, CYSLTR2, DIO1, DPH3B, ERO1LB,
FN1, GABRB2, IGFBP6, IKZF3, KIT, KRT19, LIFR, L1PH, MACC1, MAFG,
MPZL2, MT1F, MT1H, MT1P2, NELL2, ODZ1, RAG2, ROS1, SERPINA1,
SERPINA2, SLC34A2, TCFL5, TIMP1, TPO, ZMAT4, ADAMTS9, ALDH1B1,
ALDH6A1, ANO5, APOO, C10orf72, C11orf74, C14orf45, C2orf40, C4A,
C4B, C5orf28, C6orf174, CAMK2N1, CCDC121, CCND1, CDH3, CITED1,
COPZ2, CPNE3, CRABP2, CSGALNACT1, DAPK2, DLC1, ECE1, EIF2B2, EMP2,
ERBB2, FAM82B, FIBIN, FLJ42258, FRMD3, HEY2, HRASLS, ID3, IGF2BP2,
IGSF1, IKZF2, ITGA9, KIAA0408, KIAA1305, LMO3, MATN2, MDK, MET,
METTL7B, MFGE8, MGC2889, MIS12, NAV2, NCAM1, NIPSNAP3A, NIPSNAP3B,
NODI, NTAN1, NUCB2, NUPR1, PCMTD1, PIGN, PLAG1, PSAT1, PXDNL,
QTRT1, RG9MTD2, RXRG, SDC4, SLC35D2, SLC7A11, SMAD9, SPRY1, STEAP2,
TASP1, TCF7L2, TMEM171, TNFRSF11B, TNFRSF12A, TRPC5, TXNL1, WDR72,
YIPF1, ZCCHC12, ZCCHC16.
[0231] (ii) List 5: FC subtype, n=102:
[0232] TCID-3124344, ABCC3, ANGPTL1, AVPR1A, C8orf39, CD2, CD36,
CD48, CD52, CKB, CLDN1, CLDN16, CRABP1, CXCL9, DI01, DLG2,
DNASE1L3, DPH3B, DYNLT1, EIF4H, ERO1LB, F8, FABP4, FBLN5, FLJ26056,
FXYD6, FYB, GLIS3, GULP1, GZMA, GZMK, HK1, HLA-DPB1, IFITM1,
IGFBP7, IGJ, IGK@, IGKC, IGKV1-5, IGKV3-15, IGKV3-20, IGKV3D-11,
IGKV3D-15, IPCEF1, KHDRBS2, KLHDC8A, KLRC4, KLRK1, LAMB1, LCP1,
LIFR, LOC100130100, LOC100131869, LOC26080, LOC284861, LOC439911,
LOC440871, LOC650405, LOC652493, LOC652694, LOC653264, LOC728212,
LOC729461, LYZ, MAGOH2, MAPK4, MT1F, MT1H, MT1P2, NEB, ODZ1,
PLA2G7, POR, PRMT6, PSMB2, PTPRC, RAG2, RNASE1, RNASET2, RPL9P11,
RRAS2, RRBP1, RYR2, SCP2, SERGEF, SGPP2, SH3BGR, SLC25A33, SLC26A4,
SQLE, STK32A, TCFL5, TFCP2L1, TIAM2, TIMP3, TMEM220, TPO, TRPC6,
TSSC1, VAMP1, ZFPM2, ZNF487.
[0233] (iii) List 6: LCT subtype, n=140:
[0234] ADAMTS9, AIM2, APOBEC3F, APOBEC3G, ARHGAP19, ATP13A4, BAG3,
BCL2A1, BIRC5, BLNK, C10orf72, C11orf72, C12orf35, C4orf7,
C6orf168, CALCA, CARD17, CARD8, CASP1, CCL19, CCND1, CD180, CD2,
CD3D, CD48, CD52, CD79A, CD96, CEP110, CHGB, CLDN16, CLEC2B, CNN2,
COL12A1, CR2, CXCL13, CXCL9, CYTH1, DENND4A, DNAJB14, DOCK8, DPYD,
DUOX1, DUOX2, DUOXA1, DUOXA2, DUSP6, DYNC112, EGF, EPDR1, EPR1,
EPS8, ETS1, FLJ42258, FYB, GABBR2, GABRB2, GALNT7, GBP5, GIMAP2,
GIMAP5, GIMAP7, GPR155, GPR174, GTF3A, GZMA, GZMK, HIST1H3B,
HIST1H4L, HLA-DPB1, HNRNPM, IFI16, IFITM1, IFNAR2, IGF2BP2, IGJ,
IGK@, IGKC, IGKV1-5, IGKV3-15, IGKV3-20, IGKV3D-11, IGKV3D-15,
IKZF3, IL7R, ITM2A, JAK2, KBTBD8, KLHL6, KLRC4, KLRG1, KLRK1, KYNU,
LCP1, LIPH, LOC100130100, LOC100131490, LOC440871, LOC646358,
LOC650405, LOC652493, LOC652694, LONRF2, LYZ, MED13L, METTL7B,
MPZL2, MTIF3, NAV2, ND1, NFATC3, ODZ1, PAPSS2, PROS 1, PSD3, PTPRC,
PYGL, PYHIN1, RAD23B, RGS13, RIMS2, RRM2, SCG3, SLIT1, SP140,
SP140L, SPC25, ST20, ST3GAL5, STAT4, STK32A, TC2N, TLE4, TNFAIP8,
TNFRSF17, TNFSF10, TOX, UCHL5, ZEB2, ZNF143.
[0235] (iv) List 7: FVPTC subtype, n=182:
[0236] ABCC3, ADAMTS9, AIDA, ALDH1B1, ALDH6A1, ANK2, ANO5, APOL1,
APOO, AQP4, ATP13A4, BMP8A, C10orf72, C11orf72, C11orf74, C12orf35,
C14orf45, C2orf40, C4A, C4B, C5orf28, C6orf174, C8orf79, CAMK2N1,
CCDC121, CCND1, CCND2, CD36, CDH3, CITED1, CLDN1, CLDN16, CLDN4,
CLEC2B, CLU, COPZ2, CPNE3, CRABP2, CSGALNACT1, CST6, CWH43,
CYSLTR2, DAPK2, DCAF17, DIO1, DIRAS3, DLC1, DOCK9, DPH3B, DUOX1,
DUOX2, DUOXA1, DUOXA2, DUSP6, ECE1, EIF2B2, EMP2, ERBB2, ERO1LB,
ESRRG, FABP4, FAM82B, FAT4, FIBIN, FLJ42258, FN1, FRMD3, GABBR2,
GABRB2, GIMAP2; GIMAP7, GPR155, GPR98, GTF3A, GZMA, GZMK, HEY2,
HRASLS, ID3, IGF2BP2, IGFBP6, IGSF1, IKZF2, IKZF3, ITGA9, JAK2,
KIAA0284, KIAA0408, KIAAl217, KIAA1305, KIT, KLRC4, KLRK1, KRT19,
LGALS3, LIFR, LIPH, LMO3, LOC100131490, LOC100131993, LRP1B, LRP2,
MACC1, MAFG, MAPK6, MATN2, MDK, MET, METTL7B, MFGE8, MGC2889,
MIS12, MPPED2, MPZL2, MT1F, MT1G, MT1H, MT1P2, MTIF3, NAV2, NCAM1,
NELL2, NFATC3, NIPSNAP3A, NIPSNAP3B, NODI, NRCAM, NTAN1, NUCB2,
NUPR1, ODZ1, PCMTD1, PDE5A, PIGN, PKHD1L1, PLA2R1, PLAG1, PLSCR4,
PRINS, PSAT1, PXDNL, QTRT1, RAG2, RCBTB2, RG9MTD2, ROS1, RPS6KA6,
RXRG, SALL1, SCG5, SDC4, SERPINA1, SERPINA2, SLC26A4, SLC34A2,
SLC35D2, SLC7A11, SMAD9, SPRY1, ST3GAL5, STEAP2, STK32A, TASP1,
TCF7L2, TCFL5, TIMP1, TMEM171, TMEM215, TNFAIP8, TNFRSF11B,
TNFRSF12A, TNFSF10, TPO, TRPC5, TXNL1, UCHL5, WDR72, YIPF1,
ZCCHC12, ZCCHC16, ZMAT4, ZYG11A.
[0237] (v) List 8: PTC subtype, n=604:
[0238] TCID-3153400, TCID-3749600, ABCC3, ABTB2, ACBD7, ACSL1,
ACTA2, ADAMTS5, ADAMTS9, ADK, AGR2, AHNAK2, AHR, AIDA, AK1, ALAS2,
ALDH1A3, ALOX5, AMIG02, AMOT, ANK2, ANXA1, ANXA2, ANXA2P1, ANXA3,
AOAH, AP3S1, APOL1, AQP9, ARHGAP24, ARL13B, ARL4A, ARMCX3, ARMCX6,
ARNTL, ASAP2, ATIC, ATP13A4, ATP13A4, B3GNT3, BCL9, BBLHE40,
BHLHE41, BMP8A, BTBD11, BTG3, C11orf72, C11orf80, C12orf49,
C16orf45, C19orf33, C1orf115, C1orf116, C2, C2orf40, C3, C4A, C4B,
C4orf34, C6orf168, C6orf174, C7orf62, C8orf4, C8orf79, CA11, CADM1,
CAMK2N1, CAND1, CAV1, CAV2, CCDC109B, CCDC121, CCDC148, CCDC80,
CCL13, CCND1, CCND2, CD151, CD200, CD36, CDCP1, CDH11, CDH3, CDH6,
CDK2, CDKL2, CD01, CDON, CDR1, CFB, CFH, CFHR1, CFI, CHAF1B, CHD4,
CHI3L1, CITED1, CKS2, CLC, CLDN1, CLDN10, CLDN16, CLDN4, CLDN7,
CLEC4E, CLU, CNN3, COL1A1, CP, CRABP1, CRABP2, CSF3R, CST6,
CTNNAL1, CTNNB1, CTSC, CTSH, CTTN, CXCL1, CXCL14, CXCL17, CXCL2,
CXCL3, CXorf18, CXorf27, CYP1B1, CYSLTR2, DAPK2, DCBLD2, DCUN1D3,
DDAH1, DDB2, DDX52, DGKH, DGKI, DHRS1, DHRS3, DIO1, DIRAS3, DLC1,
DOCK9, DPP4, DPYSL3, DSG2, DSP, DST, DUSP4, DUSP5, DUSP6, DZIP1,
ECE1, EDNRB, EGFR, EHBP1, EHD2, EHF, ELK3, ELMO1, EMP2, EMR3, ENAH,
ENDOD1, EPB41, EPHA4, EPHX4, EPS8, ERBB3, ERI2, ERP27, ESRRG,
ETNK2, ETV1, ETV5, F2RL2, FAAH2, FABP4, FAM111A, FAM111B, FAM164A,
FAM176A, FAM20A, FAM55C, FAM84B, FBX02, FBX021, FCN1, FCN2, FGF2,
FGFR10P2, F1120184, FLJ32810, FLJ42258, FLRT3, FN1, FPR1, FPR2,
FRMD3, FZD4, FZD6, FZD7, GOS2, GABBR2, GABRB2, GADD45A, GALE,
GALNT12, GALNT3, GALNT7, GBP1, GBP3, GGCT, GLDN, GNG1.2, GOLT1A,
GPAM, GPR110, GPR110, GPR125, GPR98, GPRC5B, GRAMD3, GSN, GYPB,
GYPC, GYPE, HEMGN, HEY2, HIGD1A, HIST1H1A, HLA-DQB2, HLF, HMGA2,
HPN, HSPH1, ICAM1, IGF2BP2, IGFBP5, IGFBP6, IGSF1, IKZF3, IL1RAP,
IL1RL1, IL8RA, IL8RB, IL8RB, IL8RBP, IL8RBP, IMPDH2, INPP5F,
IPCEF1, IQGAP2, ITGA2, ITGA3, ITGA9, ITGB1, ITGB6, ITGB8, ITPR1,
JUB, KAL1, KATNAL2, KCNK5, KCNQ3, KCTD14, KDELC1, KDELR3, KHDRBS2,
KIAA0284, KIAA0408, KIAAl217, KIT, KLF8, KLK10, KLK7, KRT18, KRT19,
LAMB3, LAMC1, LAMC2, LCA5, LCMT1, LCN2, LDOC1, LEMD1, LGALS3,
LILRA1, LILRB1, LIMA1, LINGO2, LIPH, LMO3, LMO4, LOC100124692,
LOC100127974, LOC100129112, LOC100129115, LOC100129171,
LOC100129961, LOC100130248, LOC100131102, LOC100131490,
LOC100131938, LOC100132338, LOC100132764, LOC283508, LOC440434,
L00554202, LOC643454, LOC648149, LOC653354, LOC653498, LOC730031,
LONRF2, LOX, LPAR5, LPL, LRP1B, LRP2, LRRC69, LRRN1, LUM, LYRM1,
MACC1, MAFG, MAMLD1, MAP2, MAPK6, MATN2, MBOAT2, MCM4, MCM7, MDK,
MED13, MET, METTL7B, MEX3C, MFGE8, MGAM, MGAT4C, MGST1, MLLT4,
MMP16, MMP16, MNDA, MORC4, MPPED2, MPZL2, MRPL14, MT1F, MT1G, MT1H,
MT1M, MT1P2, MT1P3, MTHFD1L, MUC1, MUC15, MVP, MXRA5, MYEF2, MYH10,
MYO1B, MYO1D, MY06, NAB2, NAE1, NAG20, NCKAP1, NDFIP2, NEDD4L,
NELL2, NEXN, NFE2, NFIB, NFKBIZ, NIPAL3, NOD1, NPC2, NPEPPS, NPY1R,
NRCAM, NRIP1, NRP2, NT5E, NUDT6, OCIAD2, OCR1, ODZ1, OSGEP, OSMR,
P2RY13, P4HA2, PAM, PARP14, PARP4, PARVA, PBX1, PDE5A, PDE9A,
PDGFRL, PDLIM1, PDLIM4, PDZRN4, PEG10, PERP, PHEX, PHF16, PHLDB2,
PHYHIP, PKHD1L1, PKP4, PLA2G16, PLA2R1, PLAG1, PLAU, PLCD3,
PLEKHA4, PLEKHA5, PLK2, PLP2, PLS3, PLXNC1, PMEPA1, PON2, PPARGC1A,
PPBP, PPL, PPP1R14C, PRICKLE1, PRINS, PROK2, PROS1, PRR15, PRRG1,
PRSS23, PSD3, PTPN14, PTPRE, PTPRF, PTPRG, PTPRK, PTRF, QTRT1,
RAB25, RAB27A, RAB34, RAD23B, RAG2, RAI2, RAPGEF5, RARG, RASA1,
RASD2, RBBP7, RBBP8, RBMS2, RCE1, RDH5, RGS18, RGS2, RHOU, RND3,
ROS1, RPL39L, RPRD1A, RPS6KA6, RRAS, RUNX1, RUNX2, RXRG, S100Al2,
S100A14, S100A16, S100A8, S100A9, SALL1, SAV1, SC4MOL, SCARA3,
SCARNA11, SCEL, SCG5, SCNN1A, SCRN1, SDC4, SEH1L, SEL1L3, SELL,
SEMA3D, SEPT11, SERINC2, SERPINA1, SERPINA2, SERPINE2, SERPING1,
SFN, SFTPB, SGCB, SGCE, SGEF, SGMS2, SH2D4A, SH3PXD2A, SIRPA,
SIRPB1, SLA, SLC12A2, SLC16A4, SLC17A5, SLC24A5, SLC26A4, SLC26A7,
SLC27A2, SLC27A6, SLC34A2, SLC35F2, SLC39A6, SLC4A4, SLC5A8,
SLC7A2, SLIT2, SLPI, SMOC2, SMURF2, SNCA, SNX1, SNX22, SNX7,
SORBS2, SPATS2, SPATS2L, SPINT1, SPRED2, SPRY1, SPRY2, SRL, SSPN,
ST3GAL5, STK32A, SULF1, SYNE1, SYT14, SYTL5, TACSTD2, TBC1D3F,
TDRKH, TEAD1, TEAD1, TFCP2L1, TFF3, TGFA, TGFB2, TGFBR1, TIMP 1,
TIPARP, TJP1, TJP2, TLCD1, TLR8, TM4SF1, TM4SF4, TM7SF4, TMEM100,
TMEM117, TMEM133, TMEM163, TMEM215, TMEM90A, TMEM98, TMPRSS4,
TMSB10, TNC, TNFRSFI2A, TNFSF15, TOMM34, TPD52L1, TP0, TR1P10,
TRPC5, TSC22D1, TSPAN13, TSPAN6, TUBB1, TUBB6, TULP3, TUSC3, TYMS,
VNN2, VNN3, WDR40A, WDR54, WNT5A, XKRX, XPR1, YIF1B, YTHDC2,
ZBTB33, ZCCHC12, ZCCHC16, ZFP36L1, ZMAT3, ZMAT4, ZNF643,
ZNF804B.
[0239] (vi) List 9: NHP subtype, n=653:
[0240] TCID-3153400, TCID-3749600, ABTB2, ACBD7, ACSL1, ACTA2,
ADAMTS5, ADAMTS9, ADK, AGR2, AHNAK2, AHR, AIDA, AK1, AKR1C3, ALAS2,
ALDH1A3, AMIG02, AMOT, ANK2, ANO5, ANXA1, ANXA3, ANXA6, AOAH,
AP3S1, APOO, AQP4, AQP9, ARHGAP24, ARL13B, ARL4A, ARMCX3, ARMCX6,
ARNTL, ARSG, ASAP2, ATIC, ATP13A4, ATP6VOD2, B3GNT3, BCL9, BHLHE40,
BHLHE41, BMP8A, BTBD11, BTG3, C10orf72, C11orf72, C11orf74,
C11orf80, C12orf49, C16orf45, C19orf33, C1orf115, C1orfll6, C2,
C22orf9, C2orf40, C3, C4A, C4B, C4orf34, C5orf28, C6orf168,
C6orf174, C7orf62, C8orf4, C8orf79, C9orf68, CA11, CADM1, CALCA,
CAMK2N1, CAND1, CASC5, CAV1, CAV2, CCDC121, CCDC148, CCDC80, CCL13,
CCND1, CCND1, CCND2, CD151, CD200, CD36, CDCP1, CDH11, CDH3, CDH6,
CDK2, CDKL2, CDO1, CDON, CDR1, CEP55, CFB, CFH, CFHR1, CFI, CHAF1B,
CHD4, CITED1, CKS2, CLC, CLDN1, CLDN10, CLDN16, CLDN4, CLDN7,
CLEC4E, CLU, CNN3, COL1A1, COPZ2, CP, CPE, CRABP1, CRABP2, CSF3R,
CST6, CTNNAL1, CTNNB1, CTSH, CTTN, CWH43, CXCL1, CXCL14, CXCL17,
CXCL2, CXCL3, CXorf18, CXorf27, CYP24A1, CYP27A1, CYSLTR2, DAPK2,
DCAF17, DCBLD2, DCUN1D3, DDAH1, DDB2, DDX52, DGKH, DGKI, DHRS1,
DHRS3, DI01, DIRAS3, DLC1, DLGAP5, DOCK9, DPP4, DPYSL3, DSG2, DSP,
DST, DUOX1, DUOX2, DUOXA1, DUOXA2, DUSP4, DUSP5, DUSP6, DZIP1,
ECE1, EDNRB, EGFR, EHBP1, EHD2, EHF, ELKS, ELMO1, EMP2, EMR3, ENAH,
ENDOD1, EPB41, EPHA4, EPHX4, EPS8, ERBB3, ERI2, ERP27, ESRRG,
ETNK2, ETV1, ETV5, F2RL2, FAAH2, FABP4, FAM111A, FAM111B, FAM164A,
FAM176A, FAM20A, FAM84B, FAT4, FBX02, FBX021, FCN1, FCN2, FGF2,
FGFR10P2, FLJ20184, FLJ32810, FLJ42258, FLJ42258, FLRT3, FN1, FPR1,
FPR2, FREM2, FRMD3, FXYD6, FZD4, FZD6, FZD7, GOS2, GABBR2, GABRB2,
GADD45A, GALE, GALNT12, GALNT3, GALNT7, GBE1, GBP1, GBP3, GGCT,
GLA, GLDN, GNG12, GOLT1A, GPR110, GPR110, GPR125, GPR98, GPRC5B,
GRAMD3, GSN, GYPB, GYPC, GYPE, HEMGN, HEY2, HIST1H1A, HLA-DQB2,
HMGA2, HMMR, HPN, HSD17B6, HSPH1, ICAM1, IGFBP5, IGFBP6, IGSF1,
IKZF2, IL1RL1, IL2RA, IL8, IL8RA, IL8RB, IL8RB, IL8RBP, IL8RBP,
IMPDH2, INPP5F, EPCEF1, IQGAP2, ITGA2, ITGA3, ITGA9, ITGB1, ITGB6,
ITGB8, ITPR1, JUB, KAL1, KCNK5, KCNQ3, KCTD14, KDELC1, KDELR3,
KHDRBS2, KIAA0284, KIAA0408, KIAAl217, KIF'11, KIT, KLF8, KLK10,
KLK7, KRT18, KRT19, LAMB3, LAMC1, LAMC2, LCA5, LCMT1, LCN2, LDOC1,
LEMD1, LGALS3, LILRA1, LILRB1, LIMA1, LINGO2, LIPH, LMO3, LMO4,
LOC100124692, LOC100127974, LOC100129112, LOC100129115,
LOC100129171, LOC100129961, LOC100130248, LOC100131102,
LOC100131490, LOC100131938, LOC100131993, LOC100132338,
LOC100132764, LOC283508, LOC440434, L00554202, LOC643454,
LOC648149, LOC653354, LOC653498, LOC730031, LONRF2, LOX, LPAR1,
LPAR5, LPL, LRP1B, LRP2, LRRC69, LRRN1, LUM, LYRM1, MACC1, MAFG,
MAMLD1, MAP2, MAPK6, MATN2, MBOAT2, MCM4, MCM7, MDK, ME1, MED13,
MELK, MET, METTL7B, MEX3C, MFGE8, MGAM, MGAT1, MGAT4C, MGST1,
MKI67, MLLT4, MMP16, MMP16, MNDA, MORC4, MPPED2, MPZL2, MRPL14,
MT1F, MT1G, MT1H, MT1M, MT1P2, MT1P3, MTHFD1L, MUC1, MUC15, MVP,
MXRA5, MYEF2, MYH10, MYO1B, MYO1D, MYO5A, MYO6, NAB2, NAE1, NAG20,
NAV2, NCKAP1, NDC80, NDFIP2, NEDD4L, NELL2, NEXN, NFE2, NFIB,
NIPAL3, NOD1, NPC2, NPEPPS, NPL, NPY1R, NRCAM, NRIP1, NRP2, NT5E,
NUCB2, NUDT6, NUSAP1, OCIAD2, OCRI, ODZ1, ORAOV1, OSBPL1A, OSGEP,
OSMR, P2RY13, P4HA2, PAM, PAPSS2, PARP4, PARVA, PBX1, PDE5A, PDE9A,
PDGFRL, PDLIM1, PDLIM4, PDZRN4, PEG10, PERP, PGCP, PHEX, PHF16,
PHLDB2, PHYHIP, PKHD1L1, PKP4, PLA2G16, PLA2G7, PLA2R1, PLAG1,
PLAU, PLCD3, PLCL1, PLEKHA4, PLEKHA5, PLK2, PLS3, PLSCR4, PMEPA1,
PON2, PPARGC1A, PPBP, PPL, PPP1R14C, PRCP, PRICKLE1, PRINS, PROK2,
PROS1, PRR15, PRRG1, PRSS23, PSD3, PSD3, PTPN14, PTPRE, PTPRF,
PTPRG, PTPRK, PTRF, QTRT1, RAB25, RAB27A, RAB32, RAB34, RAD23B,
RAG2, RAI2, RAPGEF5, RARG, RASA1, RASD2, RBBP7, RBBP8, RBMS2,
RCBTB2, RCE1, RDH5, RGS18, RGS2, RHOU, RND3, ROS1, RPL39L, RPRD1A,
RPS6KA6, RRAS, RXRG, S100Al2, S100A14, S100A16, S100A8, S100A9,
SALL1, SAV1, SC4MOL, SCARA3, SCARNA11, SCEL, SCG5, SCNN1A, SCRN1,
SDC4, SEH1L, SELL, SEMA3C, SEMA3D, SEPT11, SERINC2, SERPINA1,
SERPINA2, SERPINE2, SERPING1, SFN, SFTPB, SGCB, SGCE, SGEF, SGMS2,
SH2D4A, SH3PXD2A, SIRPA, SIRPB1, SLA, SLC12A2, SLC16A4, SLC16A6,
SLC17A5, SLC24A5, SLC26A4, SLC26A7, SLC27A2, SLC27A6, SLC34A2,
SLC35F2, SLC39A6, SLC4A4, SLC5A8, SLC7A11, SLC7A2, SLIT2, SLPI,
SMOC2, SMURF2, SNCA, SNX1, SNX22, SNX7, SOAT1, SORBS2, SPATS2,
SPATS2L, SPINT1, SPRED2, SPRY1, SPRY2, SRL, SSPN, ST3GAL5, STK32A,
STXBP6, SULF1, SYNE1, SYT14, SYTL5, TACSTD2, TBC1D3F, TDRKH, TEAD1,
TEAD1, TFCP2L1, TFF3, TFPI, TGFA, TGFB2, TGFBR1, TIMP1, TIPARP,
TJP1, TJP2, TLCD1, TLR8, TM4SF1, TM4SF4, TM7SF4, TMEM100, TMEM117,
TMEM133, TMEM163, TMEM171, TMEM215, TMEM90A, TMEM98, TMPRSS4, TNC,
TNFRSFI2A, TNFSF15, TOMM34, TPD52L1, TPO, TPX2, TRIP10, TRPC5,
TSC22D1, TSPAN13, TSPAN6, TUBB1, TUBB6, TULP3, TUSC3, TXNRD1, TYMS,
UCHL5, VNN1, VNN2, VNN3, WDR40A, WDR54, WIPI1, WNT5A, XKRX, XPR1,
YIF1B, YTHDC2, ZBTB33, ZCCHC12, ZCCHC16, ZFP36L1, ZMAT3, ZMAT4,
ZNF643, ZNF804B, ZYG11A.
[0241] (vii) List 10: MTC subtype, n=48:
[0242] ANXA3, ATP13A4, BLNK, C10orf131, C6orf174, C8orf79, CALCA,
CHGB, CP, CPE, DSG2, FREM2, GPR98, IGJ, IYD, KIAA0408,
LOC100129171, LPCAT2, LRRC69, MACC1, MAPK6, MGAT4C, MGST1, MMP16,
MT1G, MT1H, MT1M, MT1P2, MT1P3, MUC15, MYEF2, NT5E, PKHD1L1, PLS3,
RBMS2, RIMS2, SCG3, SEMA3D, SLA, SLC24A5, SMOC2, SULF1, TOX, TSHZ2,
TSPAN6, WDR72, ZFP36L1, ZNF208.
[0243] (viii) List 11: HC subtype, n=65:
[0244] AIM2, APOBEC3F, APOBEC3G, ARHGAP19, BAG3, BCL2A1, BMP8A,
C9orf68, CARD17, CARD8, CASP1, CD3D, CD96, CEP110, CLEC2B, CNN2,
CPE, CYTH1, DENND4A, DNAJB14, DOCK8, DPYD, DUOX1, DUOX2, DYNC112,
EGF, EPDR1, ETS1, GBP5, GIMAP2, GIMAP5, GIMAP7, GPR174, GZMK,
HNRNPM, HSD17B6, IFI16, IFNAR2, IKZF3, 1L7R, ITM2A, JAK2, KCNAB1,
KHDRBS2, KLRC4, KLRG1, KLRK1, KYNU, LOC646358, MED13L, ND1, NFATC3,
PAPSS2, PGCP, PTPRC, PYHIN1, SLIT1, SP140, SP140L, ST20, STAT4,
TC2N, TLE4, ZEB2, ZNF143.
[0245] (ix) List 12: HA subtype, n=24:
[0246] BCL2, CADM1, CAV1, CRABP1, CTNNB1, CYTH1, DIRAS3, IFITM1,
IGFBP5, IGFBP6, LOX, MAP2, MATN2, MET, MKI67, MYO1B, ND1, NUCB2,
SCG5, SCNN1A, SEL1L3, SGCE, TNFSF10, TRPC6.
[0247] (x) List 13: ATC Subtype, n=12:
[0248] CASC5, CEP55, COL12A1, DLGAP5, HMMR, KIF11, MELK, MKI67,
NDC80, NUSAP1, PYGL, TPX2.
[0249] Dominant gene ontology of top 948 thyroid biomarkers are
listed below:
[0250] List 14: Angiogenesis, n=23
[0251] ACTA2, ANXA2, ARHGAP24, CALCA, CAV1, CITED1, COL1A1, CXCL17,
EGF, ELK3, IL8, LOX, PLCD3, PROK2, RASA1, SEMA3C, TCF7L2, TGFA,
TGFB2, TIPARP, TNFRSF12A, ZFP36L1, ZFPM2.
[0252] List 15: Apoptosis, n=43
[0253] AHR, ANXA1, BAG3, BCL2, BCL2A1, MRCS, C8orf4, CADM1, CD2,
CLU, CTNNB1, DAPK2, DLC1, DNASE1L3, ECE1, ELMO1, FAM176A, FGF2,
GADD45A, GULP1, GZMA, HIPK2, IL2RA, IL8RB, JAK2, NCKAPI, NOD1,
NUPR1, PEG10, PERP, PROK2, RYR2, SLC5A8, STK17B, SULF1, TCF7L2,
TGFB2, TNFAIP8, TNFRSF11B, TNFRSF12A, TNFSF10, VNN1, ZMAT3.
[0254] List 16: Cell Cycle, Transcription Factors, n=184
[0255] AEBP1, AHR, AK1, ANXA1, APOBEC3F, APOBEC3G, ARHGAP24, ARNTL,
ATM, BCL2, BHLHE40, BHLHE41, MRCS, BMP1, BMP8A, CADM1, CAND1,
CARD8, CASP1, CCND1, CCND2, CDK2, CEP110, CEP55, CHAF1B, CHD4,
CITED1, CKS2, CLU, CRABP2, CSGALNACT1, CTNNB1, CXCL1, CXCL17,
DENND4A, DLGAP5, DST, DZIP1, EGF, EHF, E1F2B2, EIF4H, ELK3, EMP2,
EPS8, ERBB2, ERBB3, ESRRG, ETS1, ETV1, ETV4, ETV5, FABP4, FGF2,
GOS2, GADD45A, GLDN, GLIS3, GTF3A, HEMGN, HEY2, HIPK2, HLF, HMGA2,
HPN, ID3, IFI16, IFNAR2, IGSF1, IKZF2, IKZF3, IKZF4, IL2RA, IL8,
ITPR1, JAK2, JUB, KHDRBS2, KIF11, KLF8, KLK10, KRT18, LGALS3, LIFR,
LMO3, LMO4, LRP2, LTBP2, LTBP3, MACC1, MAFG, MAMLD1, MAPK4, MAPK6,
MCM4, MCMI, MDK, MED13, MED13L, MIS12, MKI67, MLLT3, MNDA, MTIF3,
MYH10, NAB2, NAE1, NDC80, NFATC3, NFE2, NFIB, NFKBIZ, NOD1, NPAS3,
NPAT, NRIP1, NRP2, NUDT6, NUPR1, NUSAP1, OSMR, PARD6B, PARP14,
PARP4, PBX1, PDLIM1, PEG10, PIAS3, PLAG1, POU2F3, PPARGC1A, PPBP,
PRMT6, PROK2, PTRF, PYHIN1, RARG, RBBP7, RBBP8, RGS2, RHOH, RRM2,
RUNX1, RUNX2, RXRG, SALL1, SEMA3D, SERPINE2, SLIT1, SLIT2, SMAD9,
SMURF2, SP140, SPC25, SPOCK1, STAT4, SYNE1, TACSTD2, TCF7L2, TCFL5,
TEAD1, TFCP2L1, TGFA, TGFB2, TGFBR1, TLE4, TNFAIP8, TNFRSF12A,
TNFRSF17, TPX2, TSC22D1, TSHZ2, TULP3, TYMS, WNT5A, ZBTB33,
ZCCHC12, ZEB2, ZFP36L1, ZFPM2, ZNF143, ZNF208, ZNF487, ZNF643.
[0256] List 17: Cell Membrane, n=410
[0257] ABCC3, ABCD2, ACSL1, ADAMTS5, ADAMTS9, ADORA1, AFAP1, AK1,
ALOX5, AMIG02, ANK2, ANO5, AP3S1, APOL1, APOO, AQP4, AQP9, ARMCX3,
ARMCX6, ASAP2, ATP13A4, ATP6VOD.sub.2, ATP8A1, AVPR1A, B3GNT3,
BCL2, BLNK, BTBD11, C10orf72, C17orf87, C1orf115, C4orf34, C5orf28,
C6orf174, CADM1, CAMK2N1, CAV1, CAV2, CCDC109B, CD151, CD180, CD2,
CD200, CD36, CD3D, CD48, CD48, CD52, CD69, CD79A, CD96, CDCP1,
CDH11, CDH3, CDH6, CDON, CFB, CFI, CHI3L1, CLDN1, CLDN10, CLDN16,
CLDN4, CLDN7, CLEC2B, CLEC4E, COL12A1, COL1A1, COPZ2, CP, CPE, CR2,
CSF3R, CSGALNACT1, CTNNAL1, CTNNB1, CWH43, CYP1B1, CYP27A1, CYP4B1,
CYSLTR1, CYSLTR2, CYTH1, DCAF17, DCBLD2, DHRS3, DI01, DIRAS3, DLG2,
DLG4, DNAJB14, DOCK9, DPP4, DPYSL3, DSG2, DUOX1, DUOX2, DUOXA1,
DUOXA2, ECE1, EDNRB, EFEMP1, EGF, EGFR, EHBP1, EHD2, ELMO1, EMP2,
EMR3, ENTPD1, EPB41, EPHA4, EPHX4, ERBB2, ERBB3, ERO1LB, F2RL2, F8,
FAAH2, FAM176A, FAM84B, FAT4, FBLN5, FLRT3, FN1, FPR1, FPR2, FREM2,
FRMD3, FXYD6, FZD4, FZD6, FZD7, GABBR2, GABRB2, GALNT12, GALNT3,
GALNT7, GBP1, GBP3, GBP5, GIMAP2, GIMAP5, GJA4, GLDN, GNG12,
GOLT1A, GPAM, GPR110, GPR125, GPR155, GPR174, GPR98, GPRC5B, GYPB,
GYPC, GYPE, HIGD1A, HK1, HLA-DPB1, HNRNPM, HPN, HSD17B6, ICAM1,
IFITM1, IFNAR2, IGSF1, IL1RAP, IL1RL1, IL2RA, IL7R, IL8RA, IL8RB,
IPCEF1, ITGA2, ITGA3, ITGA4, ITGA9, ITGB1, ITGB4, ITGB6, ITGB8,
ITM2A, ITPR1, IYD, JAK2, JUB, KAL1, KCNA3, KCNAB1, KCNK5, KCNQ3,
KCTD14, KDELR3, KIAA1305, KIT, KLRB1, KLRC4, KLRG1, KLRK1, LAMB1,
LAMC1, LEMD1, LGALS3, LIFR, LILRA1, LILRB1, LINGO2, LIPH, LPAR1,
LPAR5, LPCAT2, LPL, LRP1B, LRP2, LRRN1, LRRN3, LUM, MATN2, MBOAT2,
MET, MFGE8, MGAM, MGAT1, MGAT4C, MGST1, MMP16, MPZL2, MRC2, MUC1,
MUC15, MYH10, MY06, NAE1, NCAM1, NCKAP1, ND1, NDFIP2, NIPAL3,
NPY1R, NRCAM, NRP2, NT5E, NUCB2, ODZ1, OSMR, P2RY13, PAM, PARD6B,
PARP14, PARVA, PCDH1, PCNXL2, PERP, PHEX, PHLDB2, PIGN, PKHD1L1,
PKP2, PLA2G16, PLA2R1, PLAU, PLCD3, PLEK, PLEKHA4, PLP2, PLSCR4,
PLXNC1, PMEPA1, PON2, POR, PPAP2C, PPL, PPP1R14C, PRICKLE1, PRRG1,
PSD3, PTK7, PTPRC, PTPRE, PTPRF, PTPRG, PTPRK, PTPRU, PTRF, RAB25,
RAB27A, RARG, RASA1, RASD2, RCE1, RDH5, RGS13, RHOH, RHOU, RIMS2,
RND3, ROS1, RRAS, RRAS2, RRBP1, RYR2, S100A12, SC4MOL, SCARA3,
SCEL, SCNN1A, SDC4, SDK1, SEL1L3, SELL, SEMA3C, SEMA3D, SEMA4C,
SERINC2, SERPINA1, SGCB, SGCE, SGMS2, SGPP2, SIRPA, SIRPB1,
SLC12A2, SLC16A4, SLC16A6, SLC17A5, SLC24A5, SLC25A33, SLC26A4,
SLC26A7, SLC27A2, SLC27A6, SLC34A2, SLC35D2, SLC35F2, SLC39A6,
SLC4A4, SLC5A8, SLC7A11, SLC7A2, SMURF2, SNCA, SNX1, SOAT1, SPINT1,
SPOCK1, SPRED2, SPRY1, SPRY2, SQLE, SSPN, ST3GAL5, STEAP2, STXBP6,
SYNE1, SYT14, SYTL5, TACSTD2, TFCP2L1, TFF3, TFPI, TGFA, TGFB2,
TGFBR1, TIMP1, TJP1, TJP2, TLCD1, TLR10, TLR8, TM4SF1, TM4SF4,
TM7SF4, TMEM100, TMEM117, TMEM133, TMEM156, TMEM163, TMEM171,
TMEM215, TMEM220, TMEM90A, TMEM98, TMPRSS4, TNC, TNFRSF11B,
TNFRSF12A, TNFRSF17, TNFSF10, TNFSF15, TOMM34, TPO, TRIP10, TRPC5,
TRPC6, TSPAN13, TSPAN6, TSPAN8, TULP3, TUSC3, VAMP1, VNN1, VNN2,
VNN3, WNT5A, XKRX, XPRI, YIF1B, YIPF1, ZBTB33.
[0258] List 18: Rare Membrane Components, n=55
[0259] AMOT, ANXA1, ANXA2, CALCA, CAMK2N1, CAVI, CAV2, CCDC80, CLU,
CST6, CTNNB1, CTTN, DLCI, DPP4, DSG2, DSP, DST, ENAH, GJA4, HIPK2,
ITGB1, ITGB4, JAK2, JUB, KRT19, LCP1, LRP2, MYH10, MY05A, MY06,
NEB, PARVA, PCDH1, PERP, PKP2, PKP4, PLEK, PPL, PTRF, RAB34, RASA1,
RYR2, SCEL, SGCB, SGCE, SLC27A6, SLIT1, SPRY1, SRL, SSPN, SYNE1,
TGFB2, TIAM2, TJP1, TNFRSF12A.
[0260] List 19: Cell-Cell Adhesion, n=85
[0261] AEBP1, AFAP1, AMIG02, ARHGAP24, BCL2, CADM1, CALCA, CD151,
CD2, CD36, CD96, CDH3, CDH6, CDON, CLDN1, CLDN10, COL12A1, CSF3R,
CTNNAL1, CTNNB1, DCBLD2, DLC1, DSG2, DST, EGFR, ENAH, ENTPD1,
EPDR1, F8, FAT4, FBLN5, FLRT3, FN1, FPR2, FREM2, GPR98, ICAM1,
IGFBP7, IL1RLI, ITGA2, ITGA3, ITGA4, ITGA9, ITGB1, ITGB4, ITGB6,
ITGB8, JUB, KAL1, LAMB1, LAMB3, LAMC1, LAMC2, LIMA1, MFGE8, MLLT4,
MPZL2, NCAM1, NELL2, NRCAM, NRP2, PARVA, PCDH1, PERP, PKP2, PKP4,
PLXNC1, PTK7, PTPRC, PTPRF, PTPRK, PTPRU, RHOU, RND3, SDK1, SELL,
SGCE, SIRPA, SPOCKI, SPP1, SSPN, TJP1, TNC, TNFRSF12A, VNN1.
[0262] List 20: Apical Cell Membrane, n=15
[0263] ANK2, ATP6VOD.sub.2, CTNNB1, CTNNB1, DPP4, DUOX1, ERBB2,
ERBB3, F2RL2, FZD6, LRP2, SCNN1A, SLC26A4, SLC34A2, TFF3.
[0264] List 21: Basolateral, Lateral Cell Membrane, n=28
[0265] ANK2, ANXA1, ANXA2, CADM1, CCDC80, CTNNB1, CTTN, DSP, DST,
EGFR, EPB41, ERBB2, ERBB3, FREM2, LAMB1, LAMB3, LAMC1, LAMC2, MET,
MYH10, MY06, PTPRK, SLC26A7, SMOC2, SNCA, TIMP3, TJP1, TRIP10.
[0266] List 22: Integrins, n=14
[0267] ADAMTS5, DST, FBLN5, ICAM1, ITGA2, ITGA3, ITGA4, ITGA9,
ITGB1, ITGB4, ITGB6, ITGB8, MFGE8, PLEK.
[0268] List 23: Cell Junction, n=40
[0269] AMOT, ARHGAP24, ARHGAP24, CADM1, CAMK2N1, CLDN1, CLDN10,
CLDN16, CLDN4, CLDN7, CNN2, DLG2, DLG4, DPYSL3, DSP, ENAH, GABBR2,
GABRB2, GJA4, JUB, LIMA1, MLLT4, NCKAP1, NEXN, PARD6B, PARVA,
PCDH1, PERP, PPL, PSD3, PTPRK, PTPRU, RHOU, RIMS2, SH3PXD2A, SSPN,
TGFB2, TJP1, TJP2, VAMP1.
[0270] List 24: Cell Surface, n=17
[0271] CD36, DCBLD2, DPP4, GPR98, HMMR, IL1RL1, IL8RB, ITGA4,
ITGB1, KAL1, MMP16, PTPRK, SDC4, SULF1, TGFA, TM7SF4,
TNFRSF12A.
[0272] List 25: Extracellular Space, n=156
[0273] ADAMTS5, ADAMTS9, AEBP1, AGR2, ANGPTL1, ANXA2, APOL1, APOO,
BMP1, BMP8A, C12orf49, C2, C2orf40, C3, C4A, C4B, C4orf7, CA11,
CALCA, CCDC80, CCL13, CCL19, CDCP1, CFB, CFH, CFHR1, CFI, CHGB,
CHI3L1, CLU, COL12A1, COL1A1, CP, CPE, CSF3R, CST6, CXCL1, CXCL11,
CXCL13, CXCL14, CXCL17, CXCL2, CXCL3, CXCL9, DPP4, EFEMP1, EGF,
EGFR, EMR3, ENDOD1, EPDR1, ERBB3, F8, FAM20A, FAM55C, FBLN5, FCN1,
FCN2, FGF2, FIBIN, FN1, FXYD6, GLA, GSN, GZMA, GZMK, ICAM1, IFNAR2,
IGFBP5, IGFBP6, IGFBP7, IGJ, IGKC, IGKV1-5, IGKV3-20, IGKV3D-11,
IGSF1, IL1RAP, IL1RL1, IL7R, 1L8, KAL1, KIT, KLK10, KLK7, LAMB1,
LAMBS, LAMC1, LAMC2, LCN2, LIFR, LIPH, LOC652694, LOX, LPL, LTBP2,
LTBP3, LUM, LYZ, MATN2, MDK, MFGE8, MMP16, MUC1, MUC15, MXRA5,
NCAM1, NELL2, NPC2, NUCB2, ODZ1, PAM, PDGFRL, PGCP, PLA2G7, PLA2R1,
PLAU, PON2, PPBP, PROK2, PROS1, PRRG1, PRSS23, PXDNL, RNASE1,
RNASET2, SCG3, SCG5, SEMA3C, SEMA3D, SEPP1, SERPINA1, SERPINE2,
SERPING1, SFN, SFTPB, SLIT1, SLIT2, SLPI, SMOC2, SPINT1, SPOCK1,
SPP1, SULF1, TFF3, TFPI, TGFA, TGFB2, THSD4, TIMP1, TIMP3, TNC,
TNFRSF11B, TNFSF10, TNFSF15, WNT5A.
[0274] List 26: Cytoskeleton, n=94
[0275] ACTA2, ADORA1, AFAP1, AMOT, ANK2, ANXA2, AP3S1, ARHGAP24,
ATM, ATP8A1, BCL2, BIRC5, C2orf40, CASC5, CLU, CNN2, CNN3, COL12A1,
COL1A1, COPZ2, CTNNAL1, CTNNB1, CTTN, CXCL1, DLG4, DLGAP5, DPYSL3,
DST, DYNC112, DYNLT1, EGFR, ELMO1, ENAH, EPB41, EPS8, FAM82B,
FRMD3, GPRC5B, GSN, GYPC, IGF2BP2, IQGAP2, JAK2, JUB, KATNAL2,
KIAA0284, KIF11, KRT18, LCA5, LCP1, LIMA1, LOX, LUM, MAP2, MPZL2,
MYH10, MY01B, MY01D, MY05A, MY06, NEB, NEXN, NFE2, NUSAP1, PARVA,
PDLIM1, PKP2, PLEK, PLS3, PPL, PTPN14, RHOU, RND3, S100A9, SCNN1A,
SDC4, SGCB, SGCE, SNCA, SORBS2, SPRED2, SPRY2, STK17B, SYNE1,
TGFB2, TGFBR1, TMSB10, TMSB15A, TPX2, TRIP10, TUBB1, TUBB6, VAMP1,
WIPI1.
[0276] In some embodiments, the present invention provides a method
of classifying cancer comprising the steps of: obtaining a
biological sample comprising gene expression products; determining
the expression level for one or more gene expression products of
the biological sample; and identifying the biological sample as
cancerous wherein the gene expression level is indicative of the
presence of thyroid cancer in the biological sample. This can be
done by correlating the gene expression levels with the presence of
thyroid cancer in the biological sample. In one embodiment, the
gene expression products are selected from one or more genes listed
in Table 2. In some embodiments, the method further includes
identifying the biological sample as positive for a cancer that has
metastasized to thyroid from a non-thyroid organ if there is a
difference in the gene expression levels between the biological
sample and a control sample at a specified confidence level.
[0277] Biomarkers involved in metastasis to thyroid from a
non-thyroid organ are provided. Such metastatic cancers that
metastasize to thyroid and can be diagnosed using the subject
methods of the present invention include but are not limited to
metastatic parathyroid cancer, metastatic melanoma, metastatic
renal carcinoma, metastatic breast carcinoma, and metastatic B cell
lymphoma. Exemplary biomarkers that can be used by the subject
methods to diagnose metastasis to thyroid are listed in Table
2.
TABLE-US-00003 TABLE 2 Biomarkers involved in metastasis to thyroid
Number of Type of metastasis genes Genes Top Biomarkers of Non- 73
ACADL, ATP13A4, BIRC5, BTG3, C2orf40, C7orf62, CD24, thyroid
Metastases to the CHEK1, CP, CRABP1, CXADR, CXADRP2, DIO1, DIO2,
Thyroid EPCAM, EPR1, GPX3, HSD17B6, IQCA1, IYD, KCNJ15, KCNJ16,
KRT7, LMO3, LOC100129258, LOC100130518, LPCAT2, LRRC2, LRRC69,
MAL2, MAPK6, MGAT4C, MGC9913, MT1F, MT1G, MT1H, MT1P2, MUC15, NEBL,
NPNT, NTRK2, PAR1, PCP4, PDE1A, PDE8B, PKHD1L1, PLS3, PVRL2, PVRL3,
RGN, RPL3, RRM2, SCD, SEMA3D, SH3BGRL2, SLC26A4, SLC26A7, SNRPN,
SPC25, SYT14, TBCKL, TCEAL2, TCEAL4, TG, TPO, TSHR, WDR72, ZBED2,
ZNF208, ZNF43, ZNF676, ZNF728, ZNF99 Parathyroid Metastasis to 101
TCID-2688277, ACSL3, ACTR3B, ADAM23, ADH5, Thyroid ARP11, AS3MT,
BANK1, C10orf32, C11orf41, C2orf67, C7orf62, C8orf34, CA8, CASR,
CD109, CD226, CD24, CD44, CDCA7L, CHEK1, CLDN1, CP, DIO2, DMRT2,
DNAH11, DPP4, ELOVL2, ENPEP, EPHA7, ESRRG, EYA1, FMN2, GCM2,
GPR160, GPR64, HSD17B6, ID2, ID2B, IYD, KIDINS220, KIF13B, KL,
LGI2, LMO3, LOC100131599, LOC150786, LPL, LRRC69, MAPK6, MGST1,
MT1F, MT1G, MT1H, MT1P2, MUC15, NAALADL2, NPNT, OGN, PDE8B, PEX5L,
PKHD1L1, PLA2G4A, PLCB1, PRLR, PTH, PTN, PTPRD, PTTG1, PTTG2,
PVALB, PVRL2, RAB6A, RAB6C, RAPGEF5, RARRES2, RGN, RNF217, RPE,
SACS, SEMA3D, SGK1, SLA, SLC15A1, SLC26A4, SLC26A7, SLC7A8, SPOCK3,
ST3GAL5, STXBP5, SYCP2L, TBCKL, TG, TINF2, TMEM167A, TPO, TSHR,
TTR, WDR72, YAP1, ZBED2 Melanoma Metastasis to 190 TCID-2840750,
ABCB5, AHNAK2, ALX1, ANLN, AP1S2, Thyroid APOD, ASB11, ATP13A4,
ATP1B1, ATRNL1, AZGP1, BACE2, BAMBI, BCHE, BIRC5, BRIP1, BZW1,
BZW1L1, C2orf40, C6orf218, C7orf62, CA14, CASC1, CCNB2, CD24,
CDH19, CDK2, CDKN3, CENPF, CHRNA5, CP, CRABP1, DCT, DEPDC1, DIO1,
DIO2, DLGAP5, DSCC1, DSP, EDNRB, EIF1AY, EIF4A1, ENPP1, EPCAM,
EPR1, ESRP1, FABP7, FANCI, GAS2L3, GGH, GPM6B, GPNMB, GPR19, GPX3,
GULP1, GYG2, HAS2, HEATR5A, HMCN1, HTN1, IL13RA2, IQCA1, IYD,
KCNJ15, KCNJ16, KIAA0894, KIF23, KRT7, KRTAP19-1, LGALS1, LMO3,
LOC100129171, LOC100129258, LOC100130275, LOC100130357,
LOC100130518, LOC100131821, LOC145694, LOC653653, LRP2, LRRC69,
LSAMP, LUM, MAL2, MAPK6, MGC87042, MITF, MLANA, MME, MND1, MOXD1,
MSMB, MUC15, NDC80, NEBL, NLGN1, NOX4, NPNT, NTRK2, NUDT10, NUDT11,
PAX3, PBK, PCP4, PDE3B, PDE8B, PI15, PIGA, PIR, PKHD1L1, PLP1,
PLXNC1, POLG, POMGNT1, POPDC3, POSTN, PRAME, PRAMEL, PTPRZ1, PVRL2,
PYGL, QPCT, RGN, RNF128, ROPN1, ROPN1B, RPL3, RPSA, RPSAP15,
RPSAP58, S100B, SACS, SAMD12, SCD, SEMA3C, SERPINA3, SERPINE2,
SERPINF1, SHC4, SILV, SLA, SLC16A1, SLC26A4, SLC26A7, SLC39A6,
SLC45A2, SLC5A8, SLC6A15, SNAI2, SNCA, SNORA48, SNORA67, SORBS1,
SPC25, SPP1, SPRY2, SRPX, ST3GAL6, STEAP1, STK33, TBC1D7, TBCKL,
TCEAL2, TCEAL4, TCN1, TF, TFAP2A, TG, TIMP2, TMSB15A, TMSB15B,
TNFRSF11B, TOP2A, TPO, TPX2, TRPM1, TSHR, TSPAN1, TUBB4, TYR, TYRL,
TYRP1, WDR72, ZBED2, ZNF208, ZNF43, ZNF676, ZNF728, ZNF99 Renal
Carcinoma 130 TCID-2763154, ADFP, AKR1C3, ALPK2, APOL1, ASPA,
Metastasis to Thyroid ATP13A4, ATP8A1, BHMT, BHMT2, BICC1, BIRC3,
C12orf75, C1S, C2orf40, C3, C7orf62, CA12, CDH6, CLRN3, CP, CYB5A,
DAB2, DEFB1, DIO2, EFNA5, EGLN3, EIF1AY, ENPEP, ENPP1, ENPP3,
EPCAM, ESRP1, FABP6, FABP7, FAM133B, FCGR3A, FCGR3B, FXYD2, GAS2L3,
GLYAT, GSTA1, GSTA2, GSTA5, HAVCR1, HLA-DQA1, HPS3, IGFBP3, IL20RB,
IYD, KMO, LEPREL1, LMO3, LOC100101266, LOC100129233, LOC100129518,
LOC100130232, LOC100130518, LOC100133763, LOC728640, LOX, LRRC69,
MAPK6, MGC9913, MME, MMP7, MT1G, MUC15, NEBL, NLGN1, NNMT, NPNT,
NR1H4, OPN3, OSMR, PCOLCE2, PCP4, PDE8B, PDZK1IP1, PIGA, PKHD1L1,
POSTN, PREPL, PTHLH, RPS6KA6, S100A10, SAA1, SAA2, SCD, SLC16A1,
SLC16A4, SLC17A3, SLC26A4, SLC26A7, SLC3A1, SLCO4C1, SNX10, SOD2,
SPINK1, SPP1, SYT14, TBCKL, TCEAL2, TCEAL4, TG, TMEM161B, TMEM176A,
TMEM45A, TNFAIP6, TNFSF10, TPO, TSHR, UGT1A1, UGT1A10, UGT1A3,
UGT1A4, UGT1A5, UGT1A6, UGT1A7, UGT1A8, UGT1A9, UGT2A3, UGT2B7,
VCAM1, VCAN, ZNF208, ZNF43, ZNF676, ZNF728, ZNF99 Breast Carcinoma
117 TCID-3777770, ACADL, AGR2, AGR3, ALDH1A1, ANLN, Metastasis to
Thyroid ASPM, ATP13A4, AZGP1, BIRC5, BRIP1, C10orf81, C7orf62,
C8orf79, CA2, CCNB2, CCNE2, CDC2, CDC6, CDKN3, CENPF, CHEK1, CP,
CSNK1G1, DEPDC1, DIO1, DIO2, DLGAP5, DTL, EHF, EPR1, EZH2, FAM111B,
FANCI, GALNT5, GPX3, HHEX, HPS3, IQCA1, ITGB3, IYD, KCNJ15, KCNJ16,
KIAA0101, KIF23, LMO3, LOC100129258, LOC100130518, LOC100131821,
LOC145694, LRP2, LRRC2, LRRC69, MAPK6, MELK, MGAT4C, MKI67, MND1,
MUC15, MYB, NDC80, NPY1R, NUF2, NUSAP1, PAR1, PARP8, PBK, PCP4,
PDE1A, PDE8B, PI15, PIP, PKHD1L1, POLG, PPARGC1A, PRC1, PVRL2,
PVRL3, RAD51AP1, RGN, RPL3, RRM2, SAA1, SAA2, SCD, SCGB1D2,
SCGB2A2, SEMA3C, SERPINA3, SLA, SLC26A4, SLC26A7, SNRPN, SPC25,
ST3GAL5, STK33, SULT1C2, SYCP2, SYT14, TFF1, TG, THBS1, TOP2A, TPO,
TPX2, TRPS1, TSHR, TTK, UNQ353, VTCN1, WDR72, ZBED2, ZNF208, ZNF43,
ZNF676, ZNF728, ZNF99 B cell Lymphoma 160 ACADL, AIM2, ALDH1A1,
ALG9, APP, ARHGAP29, Metastasis to Thyroid ATP13A4, ATP1B1, BCL2A1,
BIRC3, BIRC5, BTG3, C11orf74, C2orf40, C7orf62, CALCRL, CALD1,
CD180, CD24, CD48, CD52, CD53, CDH1, CNN3, COX11, CP, CPE, CR2,
CRYAB, CXADR, CXADRP2, CXorf65, DCBLD2, DIO2, DLGAP5, DSP, EAF2,
EFCAB2, ENPP1, EPCAM, EPR1, ESRP1, FABP4, FDXACB1, FNBP1L, GJA1,
GNAI1, GNG12, GPR174, GPX3, GTSF1, HCG11, IKZF3, IL2RG, IQCA1, IYD,
KCNJ16, KLHL6, LAPTM5, LCP1, LIFR, LMO3, LOC100128219,
LOC100129258, LOC100130518, LOC100131821, LOC100131938, LOC647979,
LOC729828, LPCAT2, LPHN2, LRIG3, LRMP, LRP2, LRRC6, LRRC69, MAL2,
MAOA, MAPK6, MATN2, MCOLN2, MGC9913, MGP, MKI67, MS4A1, MT1F, MT1G,
MT1H, MT1L, MT1P2, MUC15, NCKAP1, NCKAP1L, NEBL, NME5, NPNT,
NUDT12, PAR1, PBX1, PCP4, PDE8B, PDK4, PERP, PFN2, PKHD1L1, PLOD2,
PLS3, POMGNT1, PPARGC1A, PPIC, PTPRC, PTPRM, PVRL3, RASEF, RGN,
RGS13, RGS5, RHOH, RPL3, RPL37AP8, RRM2, S100A1, S100A13, SDC2,
SELL, SEMA3D, SH3BGRL2, SLC26A4, SLC26A7, SMARCA1, SNRPN, SP140,
SP140L, SPARCL1, SPC25, SPTLC3, ST20, STK17B, SYT14, TBCKL, TCEAL2,
TCEAL4, TEAD1, TG, TJP1, TLR10, TOM1L1, TOP2A, TSHR, TSPAN1,
TSPAN6, UACA, VNN2, WBP5, WDR72, ZNF208, ZNF43, ZNF676, ZNF728,
ZNF99
[0278] (viii) Classification Error Rates
[0279] In some embodiments, top thyroid biomarkers (948 genes) are
subdivided into bins (50 TCIDs per bin) to demonstrate the minimum
number of genes required to achieve an overall classification error
rate of less than 4% (FIG. 1). The original TCIDs used for
classification correspond to the Affymetrix Human Exon 1.0ST
microarray chip and each may map to more than one gene or no genes
at all (Affymetrix annotation file:
HuEx-1.sub.--0-st-v2.na29.hg18.transcript.csv). When no genes map
to a TCID the biomarker is denoted as TCID-######.
[0280] List 27: Error Rate Bin 1 (TCID 1-50 (n=50), gene symbols,
n=58)
[0281] AMIGO2, C11orf72, C11orf80, C6orf174, CAMK2N1, CDH3, CITED1,
CLDN1, CLDN16, CST6, CXorf27, DLC1, EMP2, ERBB3, FZD4, GABRB2,
GOLT1A, HEY2, HMGA2, IGFBP6, ITGA2, KCNQ3, KIAA0408, KRT19, LIPH,
LOC100129115, MACC1, MDK, MET, METTL7B, MFGE8, MPZL2, NAB2, NOD1,
NRCAM, PDE5A, PDLIM4, PHYHIP, PLAG1, PLCD3, PRICKLE1, PROS1, PRR15,
PRSS23, PTPRF, QTRT1, RCE1, RDH5, ROS1, RXRG, SDC4, SLC27A6,
SLC34A2, SYTL5, TNFRSF12A, TRPC5, TUSC3, ZCCHC12.
[0282] List 28: Error Rate Bin 2 (TCID 51-100 (n=50), gene symbols,
n=59)
[0283] AHNAK2, AIDA, AMOT, ARMCX3, BCL9, C1orf115, C1orf116, C4A,
C4B, C6orf168, CCDC121, CCND1, CDH6, CFI, CLDN10, CLU, CRABP2,
CXCL14, DOCK9, DZIP1, EDNRB, EHD2, ENDOD1, EPHA4, EPS8, ETNK2,
FAM176A, FLJ42258, HPN, ITGA3, ITGB8, KCNK5, KLK10, LAMB3, LEMD1,
LOC100129112, LOC100132338, L00554202, MAFG, MAMLD1, MED13, MYH10,
NELL2, PCNXL2, PDE9A, PLEKHA4, RAB34, RARG, SCG5, SFTPB, SLC35F2,
SLIT2, TACSTD2, TGFA, TIMP1, TMEM100, TMPRSS4, TNC, ZCCHC16.
[0284] List 29: Error Rate Bin 3 (TCID 101-150 (n=50), gene
symbols, n=52)
[0285] ABTB2, ADAMTS9, ADORA1, B3GNT3, BMP1, C19orf33, C3, CDH11,
CLIP3, COL1A1, CXCL17, CYSLTR2, DAPK2, DHRS3, DIRAS3, DPYSL3,
DUSP4, ECE1, FBXO2, FGF2, FN1, GALE, GPRC5B, GSN, IKZF4, IQGAP2,
ITGB4, KIAA0284, KLF8, KLK7, LONRF2, LPAR5, MPPED2, MUC1, NRIP1,
NUDT6, ODZ1, PAM, POU2F3, PPL, PTRF, RAPGEF5, RASD2, SCARA3, SCEL,
SEMA4C, SNX22, SPRY1, SSPN, TM4SF4, XPR1, YIF1B.
[0286] List 30: Error Rate Bin 4 (TCID 151-200 (n=50), gene
symbols, n=58)
[0287] AFAP1, ARMCX6, ARNTL, ASAP2, C2, C8orf4, CCDC148, CFB,
CHAF1B, CLDN4, DLG4, DUSP6, ELMO1, FAAH2, FAM20A, FLRT3, FRMD3,
GALNT12, GALNT7, IGFBP5, IKZF2, ISYNA1, LOC100131490, LOC648149,
LOC653354, LRP1B, MAP2, MRC2, MT1F, MT1G, MT1H, MT1P2, MYEF2,
NPAS3, PARD6B, PCDH1, PMEPA1, PPAP2C, PSD3, PTPRK, PTPRU, RAI2,
RRAS, SDK1, SERPINA1, SERPINA2, SGMS2, SLC24A5, SMURF2, SPATS2L,
SPINT1, TDRKH, TIPARP, TM4SF1, TMEM98, WNT5A, XKRX, ZMAT4.
[0288] List 31: Error Rate Bin 5 (TCID 201-250 (n=50), gene
symbols, n=53)
[0289] ABCC3, AEBP1, C16orf45, C19orf33, CA11, CCND2, CDO1, CYP4B1,
DOK4, DUSP5, ETV4, FAM111A, FN1, GABBR2, GGCT, GJA4, GPR110, HIPK2,
ITGA9, JUB, KDELR3, KIAA1217, LAMC2, LCA5, LTBP2, LTBP3, MAPK6,
NAV2, NIPAL3, OSMR, PDZRN4, PHLDB2, PIAS3, PKHD1L1, PKP2, PKP4,
PRINS, PTK7, PTPRG, RAB27A, RAD23B, RASA1, RICH2, SCRN1, SFN,
ST3GAL5, STK32A, TCERG1L, THSD4, TJP2, TM7SF4, TPO, YIF1B.
IX. Compositions
[0290] (i) Gene Expression Products and Splice Variants of the
Present Invention
[0291] Molecular profiling may also include but is not limited to
assays of the present disclosure including assays for one or more
of the following: proteins, protein expression products, DNA, DNA
polymorphisms, RNA, RNA expression products, RNA expression product
levels, or RNA expression product splice variants of the genes
provided in FIG. 2-6, 9-13, 16 or 17. In some cases, the methods of
the present invention provide for improved cancer diagnostics by
molecular profiling of about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20,
25, 30, 35, 40, 45, 50, 60, 70, 80, 90, 100, 120, 140, 160, 180,
200, 240, 280, 300, 350, 400, 450, 500, 600, 700, 800, 1000, 1500,
2000, 2500, 3000, 3500, 4000, 5000 or more DNA polymorphisms,
expression product markers, and/or alternative splice variant
markers.
[0292] In one embodiment, molecular profiling involves microarray
hybridization that is performed to determine gene expression
product levels for one or more genes selected from: FIG. 2-6, 9-13,
16 or 17. In some cases, gene expression product levels of one or
more genes from one group are compared to gene expression product
levels of one or more genes in another group or groups. As an
example only and without limitation, the expression level of gene
TPO may be compared to the expression level of gene GAPDH. In
another embodiment, gene expression levels are determined for one
or more genes involved in one or more of the following metabolic or
signaling pathways: thyroid hormone production and/or release,
protein kinase signaling pathways, lipid kinase signaling pathways,
and cyclins. In some cases, the methods of the present invention
provide for analysis of gene expression product levels and or
alternative exon usage of at least one gene of 1, 2, 3, 4, 5, 6, 7,
9, 10, 11, 12, 13, 14, or 15 or more different metabolic or
signaling pathways.
[0293] (ii) Compositions of the Present Invention
[0294] Compositions of the present disclosure are also provided
which composition comprises one or more of the following:
nucleotides (e.g. DNA or RNA) corresponding to the genes or a
portion of the genes provided in FIG. 2-6, 9-13, 16 or 17, and
nucleotides (e.g. DNA or RNA) corresponding to the complement of
the genes or a portion of the complement of the genes provided in
FIG. 2-6, 9-13, 16 or 17. The nucleotides of the present invention
can be at least about 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60,
65, 70, 75, 100, 150, 200, 250, 300, 350, or about 400 or 500
nucleotides in length. In some embodiments of the present
invention, the nucleotides can be natural or man-made derivatives
of ribonucleic acid or deoxyribonucleic acid including but not
limited to peptide nucleic acids, pyranosyl RNA, nucleosides,
methylated nucleic acid, pegylated nucleic acid, cyclic
nucleotides, and chemically modified nucleotides. In some of the
compositions of the present invention, nucleotides of the present
invention have been chemically modified to include a detectable
label. In some embodiments of the present invention the biological
sample has been chemically modified to include a label.
[0295] A further composition of the present disclosure comprises
oligonucleotides for detecting (i.e. measuring) the expression
products of the genes provided in FIG. 2-6, 9-13, 16 or 17 and
their complement. A further composition of the present disclosure
comprises oligonucleotides for detecting (i.e. measuring) the
expression products of polymorphic alleles of the genes provided in
FIG. 2-6, 9-13, 16 or 17 and their complement. Such polymorphic
alleles include but are not limited to splice site variants, single
nucleotide polymorphisms, variable number repeat polymorphisms,
insertions, deletions, and homologues. In some cases, the variant
alleles are between about 99.9% and about 70% identical to the
genes listed in FIG. 6, including about 99.75%, 99.5%, 99.25%, 99%,
97.5%, 95%, 92.5%, 90%, 85%, 80%, 75%, and about 70% identical. In
some cases, the variant alleles differ by between about 1
nucleotide and about 500 nucleotides from the genes provided in
FIG. 2-6, 9-13, 16 or 17, including about 1, 2, 3, 5, 7, 10, 15,
20, 25, 30, 35, 50, 75, 100, 150, 200, 250, 300, and about 400
nucleotides.
[0296] In some embodiments, the composition of the present
invention may be specifically selected from the top differentially
expressed gene products between benign and malignant samples, or
the top differentially spliced gene products between benign and
malignant samples, or the top differentially expressed gene
products between normal and benign or malignant samples, or the top
differentially spliced gene products between normal and benign or
malignant samples. In some cases the top differentially expressed
gene products may be selected from FIG. 2 and/or FIG. 4. In some
cases, the top differentially spliced gene products may be selected
from FIG. 3 and/or FIG. 5.
IX. Business Methods
[0297] As described herein, the term customer or potential customer
refers to individuals or entities that may utilize methods or
services of the molecular profiling business. Potential customers
for the molecular profiling methods and services described herein
include for example, patients, subjects, physicians, cytological
labs, health care providers, researchers, insurance companies,
government entities such as Medicaid, employers, or any other
entity interested in achieving more economical or effective system
for diagnosing, monitoring and treating cancer.
[0298] Such parties can utilize the molecular profiling results,
for example, to selectively indicate expensive drugs or therapeutic
interventions to patients likely to benefit the most from said
drugs or interventions, or to identify individuals who would not
benefit or may be harmed by the unnecessary use of drugs or other
therapeutic interventions.
[0299] (i) Methods of Marketing
[0300] The services of the molecular profiling business of the
present invention may be marketed to individuals concerned about
their health, physicians or other medical professionals, for
example as a method of enhancing diagnosis and care; cytological
labs, for example as a service for providing enhanced diagnosis to
a client; health care providers, insurance companies, and
government entities, for example as a method for reducing costs by
eliminating unwarranted therapeutic interventions. Methods of
marketing to potential clients, further includes marketing of
database access for researchers and physicians seeking to find new
correlations between gene expression products and diseases or
conditions.
[0301] The methods of marketing may include the use of print,
radio, television, or internet based advertisement to potential
customers. Potential customers may be marketed to through specific
media, for example, endocrinologists may be marketed to by placing
advertisements in trade magazines and medical journals including
but not limited to The Journal of the American Medical Association,
Physicians Practice, American Medical News, Consultant, Medical
Economics, Physician's Money Digest, American Family Physician,
Monthly Prescribing Reference, Physicians' Travel and Meeting
Guide, Patient Care, Cortlandt Forum, Internal Medicine News,
Hospital Physician, Family Practice Management, Internal Medicine
World Report, Women's Health in Primary Care, Family Practice News,
Physician's Weekly, Health Monitor, The Endocrinologist, Journal of
Endocrinology, The Open Endocrinology Journal, and The Journal of
Molecular Endocrinology. Marketing may also take the form of
collaborating with a medical professional to perform experiments
using the methods and services of the present invention and in some
cases publish the results or seek funding for further research. In
some cases, methods of marketing may include the use of physician
or medical professional databases such as, for example, the
American Medical Association (AMA) database, to determine contact
information.
[0302] In one embodiment methods of marketing comprises
collaborating with cytological testing laboratories to offer a
molecular profiling service to customers whose samples cannot be
unambiguously diagnosed using routine methods.
[0303] (ii) Business Methods Utilizing a Computer
[0304] The molecular profiling business may utilize one or more
computers in the methods of the present invention such as a
computer 800 as illustrated in FIG. 22. The computer 800 may be
used for managing customer and sample information such as sample or
customer tracking, database management, analyzing molecular
profiling data, analyzing cytological data, storing data, billing,
marketing, reporting results, or storing results. The computer may
include a monitor 807 or other graphical interface for displaying
data, results, billing information, marketing information (e.g.
demographics), customer information, or sample information. The
computer may also include means for data or information input 816,
815. The computer may include a processing unit 801 and fixed 803
or removable 811 media or a combination thereof. The computer may
be accessed by a user in physical proximity to the computer, for
example via a keyboard and/or mouse, or by a user 822 that does not
necessarily have access to the physical computer through a
communication medium 805 such as a modem, an interne connection, a
telephone connection, or a wired or wireless communication signal
carrier wave. In some cases, the computer may be connected to a
server 809 or other communication device for relaying information
from a user to the computer or from the computer to a user. In some
cases, the user may store data or information obtained from the
computer through a communication medium 805 on media, such as
removable media 812. It is envisioned that data relating to the
present invention can be transmitted over such networks or
connections for reception and/or review by a party. The receiving
party can be but is not limited to an individual, a health care
provider or a health care manager. In one embodiment, a
computer-readable medium includes a medium suitable for
transmission of a result of an analysis of a biological sample,
such as exosome bio-signatures. The medium can include a result
regarding an exosome bio-signature of a subject, wherein such a
result is derived using the methods described herein.
[0305] The molecular profiling business may enter sample
information into a database for the purpose of one or more of the
following: inventory tracking, assay result tracking, order
tracking, customer management, customer service, billing, and
sales. Sample information may include, but is not limited to:
customer name, unique customer identification, customer associated
medical professional, indicated assay or assays, assay results,
adequacy status, indicated adequacy tests, medical history of the
individual, preliminary diagnosis, suspected diagnosis, sample
history, insurance provider, medical provider, third party testing
center or any information suitable for storage in a database.
Sample history may include but is not limited to: age of the
sample, type of sample, method of acquisition, method of storage,
or method of transport.
[0306] The database may be accessible by a customer, medical
professional, insurance provider, third party, or any individual or
entity which the molecular profiling business grants access.
Database access may take the form of electronic communication such
as a computer or telephone. The database may be accessed through an
intermediary such as a customer service representative, business
representative, consultant, independent testing center, or medical
professional. The availability or degree of database access or
sample information, such as assay results, may change upon payment
of a fee for products and services rendered or to be rendered. The
degree of database access or sample information may be restricted
to comply with generally accepted or legal requirements for patient
or customer confidentiality. The molecular profiling company may
bill the individual, insurance provider, medical provider, or
government entity for one or more of the following: sample receipt,
sample storage, sample preparation, cytological testing, molecular
profiling, input and update of sample information into the
database, or database access.
[0307] (iii) Business Flow
[0308] FIG. 18a is a flow chart illustrating one way in which
samples might be processed by the molecular profiling business.
Samples of thyroid cells, for example, may be obtained by an
endocrinologist perhaps via fine needle aspiration 100. Samples are
subjected to routine cytological staining procedures 125. Said
routine cytological staining provides four different possible
preliminary diagnoses non-diagnostic 105, benign 110, ambiguous or
suspicious 115, or malignant 120. The molecular profiling business
may then analyze gene expression product levels as described herein
130. Said analysis of gene expression product levels, molecular
profiling, may lead to a definitive diagnosis of malignant 140 or
benign 135. In some cases only a subset of samples are analyzed by
molecular profiling such as those that provide ambiguous and
non-diagnostic results during routine cytological examination.
Alternative embodiments by which samples may be processed by the
methods of the present invention are provided in FIGS. 18b and
21.
[0309] In some cases the molecular profiling results confirms the
routine cytological test results. In other cases, the molecular
profiling results differ. In such cases, samples may be further
tested, data may be reexamined, or the molecular profiling results
or cytological assay results may be taken as the correct diagnosis.
Benign diagnoses may also include diseases or conditions that,
while not malignant cancer, may indicate further monitoring or
treatment. Similarly, malignant diagnoses may further include
diagnosis of the specific type of cancer or a specific metabolic or
signaling pathway involved in the disease or condition. Said
diagnoses, may indicate a treatment or therapeutic intervention
such as radioactive iodine ablation, surgery, thyroidectomy; or
further monitoring.
XI. Kits
[0310] The molecular profiling business may provide a kit for
obtaining a suitable sample. Said kit 203 as depicted in FIG. 19
may comprise a container 202, a means for obtaining a sample 200,
reagents for storing the sample 205, and instructions for use of
said kit. In another embodiment, the kit further comprises reagents
and materials for performing the molecular profiling analysis. In
some cases, the reagents and materials include a computer program
for analyzing the data generated by the molecular profiling
methods. In still other cases, the kit contains a means by which
the biological sample is stored and transported to a testing
facility such as the molecular profiling business or a third party
testing center.
[0311] The molecular profiling business may also provide a kit for
performing molecular profiling. Said kit may comprise a means for
extracting protein or nucleic acids including all necessary buffers
and reagents; and, a means for analyzing levels of protein or
nucleic acids including controls, and reagents. The kit may further
comprise software or a license to obtain and use software for
analysis of the data provided using the methods and compositions of
the present invention.
EXAMPLES
Example 1
Gene Expression Product Analysis of Thyroid Samples
[0312] 75 thyroid samples were examined for gene expression
analysis using the Affymetrix Human Exon 10ST array according to
manufacturer's instructions to identify genes that showed
significantly differential expression and/or alternative splicing
between malignant, benign, and normal samples. Three groups were
compared and classified according to pathological surgical
diagnosis of the tissue: benign (n=29), malignant (n=37), and
normal (n=9). The samples were prepared from surgical thyroid
tissue, snap frozen and then the RNA was prepared by standard
methods. The names and pathological classification of the 75
samples are depicted in FIG. 1.
[0313] Microarray analysis was run with XRAY version 2.69 (Biotique
Systems Inc.). Input files were normalized with full quantile
normalization (Irizarry et al. Biostatistics 2003 Apr. 4 (2):
249-64). For each input array and each probe expression value, the
array-ith percentile probe value was replaced with the average of
all array-ith percentile points. A total of 6,553,590 probes were
manipulated in the analysis. Probes with GC count less than 6 and
greater than 17 were excluded from the analysis. The expression
score for each probe-set was derived via application of
median-polish (exon RMA) to the probe scores across all input
hybridizations and probe-sets with fewer than 3 probes (that pass
all of the tests defined above) were excluded from further
analysis. Only `Core` probe-sets, corresponding to probe-sets
matching entries in the high quality databases RefSeq and Ensembl,
were analyzed. Non-expressed probes and invariant probes were also
removed from analysis for both gene level and probe set level
analyses. One-way ANOVA analysis was used to examine gene
expression at the probe set level between groups malignant and
benign.
[0314] The top 100 differentially expressed genes by gene level
analysis (i.e. those genes which showed the greatest differential
expression) were obtained from the dataset in which benign
malignant and normal thyroid samples were compared. Markers were
selected based on statistical significance after Benjamini and
Hochberg correction for false discover rate (FDR). An FDR filter
value of p<0.01 was used, followed by ranking with absolute fold
change (>1.9) calculated per maker as the highest differential
gene expression value in any group (benign malignant or normal)
divided by the lowest differential expression in the remaining two
groups. The results of this analysis are shown in FIG. 2. This
table lists three sets of calculated fold changes for any given
marker to allow comparison between the groups. The fold changes
malignant/benign, malignant/normal, and benign/normal were all
calculated by dividing the expression of one group by the
expression of another.
[0315] The top 100 alternatively spliced genes were obtained from
the dataset in which benign malignant and normal thyroid samples
were compared. Markers were selected based on statistical
significance after Benjamini and Hochberg correction for false
discovery rate (FDR). An FDR filter value of p<0.01 was used,
and markers were ranked starting with lowest p-value. The threshold
for listing a numerical value with the software used was
p<1.0E-301, any numbers having a smaller p-value were
automatically assigned a value of 0.00E+00. The results of this
analysis are shown in FIG. 3. All the markers depicted are highly
significant for alternative exon splicing.
[0316] The top 100 differentially expressed genes in the thyroid
samples from FIG. 1 by probe-set level analysis were obtained from
the dataset in which benign and malignant samples were analyzed.
Markers were selected based on significance after Benjamini and
Hochberg correction for false discovery rate (FDR). Markers were
selected based on significance after Benjamini and Hochberg
correction for false discovery rate (FDR). An FDR filter value of
p<0.01 was used, followed by ranking with absolute fold-change
(>2.0) calculated per marker as Malignant expression divided by
Benign expression. The results of this analysis are shown in FIG.
4.
[0317] The top 100 statistically significant diagnostic markers
determined by gene level analysis of the thyroid samples shown in
FIG. 1 were also compiled. Data from the comparison between benign,
malignant, and normal and from comparison between benign and
malignant datasets were used. Markers were selected based on
significance after Benjamini and Hochberg correction for false
discovery rate (FDR). An FDR filter value of p<0.01 was used,
followed by ranking with absolute fold-change (>1.6) calculated
per marker as the highest differential expression value in any
group (benign, malignant or normal) divided by the lowest
differential expression in the remaining two groups. The
fold-changes for Malignant/Benign, Malignant/Normal, and
Benign/Normal were all calculated in similar fashion by dividing
the expression of one group by the expression of another. The
results of this analysis are shown in FIG. 5.
[0318] The full list of 4918 genes identified as statistically
significantly differentially expressed, differentially spliced or
both between benign and malignant, benign and normal, or malignant
and normal samples at either the probe-set or gene level was also
compiled. Markers were selected based on statistical significance
after Benjamini and Hochberg correction for false discovery rate
(FDR), and an FDR filter value of p<0.01 was used. The results
are depicted in FIG. 6.
Example 2
Gene Expression Product Analysis of Thyroid Tissue Samples
[0319] A total of 205 thyroid tissue samples (FIG. 7) are examined
with an Affymetrix HumanExon10ST array chip to identify genes that
differ significantly in RNA expression levels between benign and
malignant samples. Samples are classified according to
post-surgical thyroid pathology: samples exhibiting follicular
adenoma (FA), lymphocytic thyroiditis (LCT), or nodular hyperplasia
(NHP) are classified as benign; samples exhibiting Hurthle cell
carcinoma (HC), follicular carcinoma (FC), follicular variant of
papillary thyroid carcinoma (FVPTC), papillary thyroid carcinoma
(PTC), medullary thyroid carcinoma (MTC), or anaplastic carcinoma
(ATC) are classified as malignant.
[0320] Affymetrix software is used to extract, normalize, and
summarize intensity data from roughly 6.5 million probes.
Approximately 280,000 core probe sets are subsequently used in
feature selection and classification. The models used are LIMMA for
feature selection and random forest and support vector machine
(SVM) for classification. Iterative rounds of training,
classification, and cross validation are performed using random
subsets of data. Top features are identified in two separate
analyses (malignant vs. benign and MTC vs. rest) using the
classification engine described above.
[0321] Markers are selected based on significance after Benjamini
and Hochberg correction for false data discovery rate (FDR). An FDR
filter of p<0.05 is used.
[0322] A malignant vs. benign comparison of thyroid tissue samples
finds 413 markers that are diagnostic for thyroid diseases or
conditions. The top 100 markers are listed in FIG. 9.
[0323] An MTC vs. the rest (i.e. non-MTC) comparison of thyroid
tissue samples finds 671 markers that are diagnostic for thyroid
diseases or conditions. The top 100 markers are listed in FIG.
10.
Example 3
Meta-analysis of Gene Expression Product Data from Thyroid
Samples
[0324] Surgical thyroid tissue samples (FIG. 7) and thyroid samples
obtained via fine needle aspiration (FIG. 8) are identified as
benign or malignant by pathological examination and then examined
by hybridization to an Affymetrix HumanExonl OST array. A
meta-analysis approach is utilized which allows the identification
of genes with repeatable features in each classification.
Affymetrix software is used to extract, normalize, and summarize
intensity data from approximately 6.5 million probes. Roughly
280,000 probe sets are used for feature selection and
classification. LIMMA is used for feature selection. Classification
is performed with random forest and SVM methods. Markers that
repeatedly appear in multiple iterative rounds of training,
classification, and cross validation of the surgical and fine
needle aspirate samples are identified and ranked. A joint set of
core features are created using the top ranked features that appear
for both the surgical and fine needle aspirate data. Markers with a
non-zero repeatability score are selected as significant. A total
of 102 markers are found to be significant and are listed in FIG.
11.
Example 4
Bayesian Analysis of Gene Expression Product Data from Thyroid
Samples
[0325] Two groups of well-characterized samples are compared in
order to identify genes that distinguish benign from malignant
nodules in the human thyroid. Samples are derived from surgical
thyroid tissue (tissue; n=205, FIG. 7) or from fine needle
aspirates (FNA; n=74, FIG. 8) and are examined by hybridization to
the HumanExonl OST microarray. Pathology labels for each distinct
thyroid subtype are coded as either benign (B) or malignant (M). A
total of 499 markers that show distinct differential expression
between benign and malignant samples are identified.
[0326] Affymetrix software is used to extract, normalize, and
summarize intensity data from approximately 6.5 million probes.
Roughly 280,000 core probe sets are subsequently used in feature
selection and classification of .about.22,000 genes. The models
used are LIMMA (for feature selection) and SVM (for classification)
respectively.
[0327] Next, we previously published molecular profile studies are
examined in order to derive the type I and type II error rates of
assigning a gene into the "benign" or "malignant" category. The
error rates are calculated based on the sample size reported in
each particular published study with an estimated fold-change value
of two. Lastly, these prior probabilities are combined with the
output of the Tissue dataset to estimate the posterior probability
of differential gene expression, and then combined with the FNA
dataset to formulate the final posterior probabilities of
differential expression (Smyth 2004). These posterior probabilities
are used to rank the genes and those that exceed a posterior
probability threshold of 0.9 are selected. A total of 499 markers
are identified as significant and the top 100 are listed in FIG.
12.
Example 5
Subtype Analysis of Gene Expression Product Data from Thyroid
Samples
[0328] Well-characterized samples are examined in order to
distinguish benign nodules from those with distinct pathology in
the human thyroid. 205 hybridizations to the HumanExon10ST
microarray are examined. Pathology labels for each distinct thyroid
subtype are used to systematically compare one group versus
another. A total of 250 mRNA markers that separate thyroid into a
wide range of pathology subtypes are identified.
[0329] A total of 205 thyroid tissue samples are examined with the
Affymetrix HumanExonl OST array chip to identify genes that differ
significantly in mRNA expression between distinct thyroid pathology
subtypes (FIG. 7). Samples classified according to post-surgical
thyroid pathology as: follicular adenoma (FA, n=22), lymphocytic
thyroiditis (LCT, n=39), nodular hyperplasia (NHP, n=24)), are all
collectively classified as benign (n=85). In contrast, samples
classfied as Hurthle cell carcinoma (HC, n=27), follicular
carcinoma (FC, n=19), follicular variant of papillary thyroid
carcinoma (FVPTC, n=21), papillary thyroid carcinoma (PTC, n=26),
medullary thyroid carcinoma (MTC, n=22), and anaplastic carcinoma
(ATC, n=5) are all collectively classified as malignant
(n=120).
[0330] Affymetrix software is used to extract, normalize, and
summarize intensity data from roughly 6.5 million probes.
Approximately 280,000 core probe sets are subsequently used in
feature selection and classification. A given benign subtype (e.g.,
NHP) set is compared against a pool of all other malignant subtypes
(e.g., NHP vs. M) next the benign subset is compared again against
each set of malignant subtypes (NHP vs. FC, NHP vs. PTC, etc). The
models used in the classification engine are LIMMA (for feature
selection), and random forest and SVM are used for classification.
Iterative rounds of training, classification, and cross-validation
are performed using random subsets of data. A joint core-set of
genes that separate distinct thyroid subtypes is created.
[0331] Markers are selected based on the set of genes that
optimizes the classifier after pair-wise classification. A total of
251 markers mapping to 250 distinct genes allow the separation of
1-3 distinct thyroid subtypes (FIG. 13).
Example 6
Differentially Expressed miRNAs Identified Via the Agilent vs
microRNA Array
[0332] Thyroid samples are hybridized to the Agilent Human v2
microRNA (miRNA) array. This array contains probes to 723 human and
76 viral miRNAs, and these are targeted using .about.15,000
probesets. A comparison between benign (B) and malignant (M)
thyroid samples is performed to identify significant differentially
expressed miRNAs. All samples are derived from clinical fine needle
aspirates (n=89, FIG. 14).
[0333] Array intensity data is extracted, normalized, and
summarized, followed by modeling using classification engine.
Briefly, the models used are LIMMA (for feature selection), and
random forest and support vector machine (SVM) are used for
classification. Iterative rounds of training, classification, and
cross-validation are performed using random subsets of data.
Although several miRNAs are differentially expressed in malignant
as compared to benign (FIG. 16), no stand-alone classifiers were
identified with this approach.
Example 7
Differentially Expressed miRNAs that are Diagnostic for Thyroid
Diseases
[0334] Thyroid nodule samples are hybridized to the Illumina Human
v2 miRNA array. This array contains probes to 1146 human miRNAs. A
comparison between benign and malignant thyroid samples is
performed to identify significant differentially expressed miRNAs.
All samples are derived from clinical FNAs (n=24, FIG. 15).
[0335] Array intensity data is extracted, normalized, and
summarized, followed by modeling using a classification engine.
Briefly, the models used are LIMMA (for feature selection), and
random forest, and support vector machine (SVM) for classification.
An additional "hot probes" method is added to the classification
engine, which in part incorporates a meta-analysis approach to the
algorithm. Iterative rounds of training, classification, and
cross-validation are performed using random subsets of data. The
"hot probes" method identifies probes that appear in every loop of
cross-validation, thereby creating a set of robust, repeatable
features. Markers are selected based on the p-value (P) of a
comparison between malignant and benign samples. A total of 145
miRNAs are identified whose differential expression is identified
as diagnostic for benign or malignant thyroid conditions (FIG.
17).
Example 8
An Exemplary Device for Molecular Profiling
[0336] The molecular profiling business of the present invention
compiles the list of 4918 genes of FIG. 6 that are differentially
expressed, differentially spliced or both between benign and
malignant, benign and normal, or malignant and normal samples at
either the probe-set or the gene level. A subset of the 4918 genes
are chosen for use in the diagnosis of biological samples by the
molecular profiling business. Compositions of short (i.e. 12-25
nucleotide long) oligonucleotides complimentary to the subset of
4918 genes chose for use by the molecular profiling business are
synthesized by standard methods known in the art and immobilized on
a solid support such as nitrocellulose, glass, a polymer, or a chip
at known positions on the solid support.
Example 9
Molecular Profiling of a Biological Sample
[0337] A biological sample is obtained by fine needle aspiration
and stored in two aliquots, one for molecular profiling and one for
cytological analysis. The aliquot of biological sample for
molecular profiling is added to lysis buffer and triturated which
results in lysing of the cells of the biological sample. Lysis
buffer is prepared as follows: For 1 ml of cDNA lysis buffer, the
following were mixed together on ice: 0.2 ml of Moloney murine
leukemia virus (MMLV) reverse transcriptase, 5.times. (Gibco-BRL),
0.76 ml of H20 (RNAse, DNAse free, Specialty Media), 5 .mu.l of
Nonidet P40 (USB), 10 .mu.l of PrimeRNase inhibitor (3'5'
Incorporated), 10 .mu.l of RNAguard (Pharmacia), and 20 .mu.l of
freshly made, 1/24 dilution of stock primer mix. The stock primer
mix, kept aliquoted at -20.degree. C., includes 10 .mu.l each of
100 mM dATP, dCTP, dGTP, dTTP solutions (12.5 mM
final)(Boehringer); 10 .mu.l of 50 OD/ml pd(T)19-24 (Pharmacia);
and 3 0 .mu.l H20.
[0338] Cell RNA is then primed with an oligo dT primer. Reverse
transcription with reverse transcriptase is then performed in
limiting conditions of time and reagents to facilitate incomplete
extension and to prepare short cDNA of between about 500 bp to
about 1000 bp. The cDNA is then tailed at the 5' end with multiple
dATP using polyA (dATP) and terminal transferase.
[0339] The cDNA is then amplified with PCR reagents using a 60mer
primer having 24(dT) at the 3' end. PCR cycling is performed at
94.degree. C. for 1 minute, then 42.degree. C. for 2 minutes and
then 72.degree. C. for 6 minutes with 10 second extension times at
each cycle. 10 cycles are performed. Then additional Taq polymerase
is added and an additional 25 cycles are performed.
[0340] cDNA is extracted in phenol-chloroform, precipitated with
ethanol and then half of the sample is frozen at -80.degree. C. as
a stock to avoid thawing and freezing the entire amount of cDNA
while analyzing it.
[0341] 5 .mu.g of PCR product is combined with 15.5 .mu.l EF sln
(Tris in Qiagen kit PCR purification), 4 .mu.l of lox One-Phor-All
buffer from Promega, and 0.5 units of DNase I. The total volume is
then held at 37.degree. C. for 14 minutes, then held at 99.degree.
C. for 15 minutes and then put on ice for 5 minutes to fragment the
PCR product into segments about 50 bp to about 100 bp in length.
The fragments are then end-labeled by combining the total volume
with 1 .mu.l of Biotin-N6-ddATP ("NEN") and 1.5 .mu.l of TdT
(terminal transferase) (15 unit/.mu.l). The total volume is then
held at 37.degree. C. for 1 hour, then held at 99.degree. C. for 15
minutes and then held on ice for 5 minutes.
[0342] The labeled and fragmented cDNA is hybridized with the
probeset of the present invention in 200 microliters of
hybridization solution containing 5-10 microgram labeled target in
1.times.MES buffer (0.1 M MES, 1.0 M NaC1, 0.01% Triton X-100, pH
6.7) and 0.1 mg/ml herring sperm DNA. The arrays used are
Affymetrix Human Exon 10ST arrays. The arrays are placed on a
rotisserie and rotated at 60 rpm for 16 hours at 45.degree. C.
Following hybridization, the arrays are washed with 6.times.SSPE-T
(0.9 M NaC1, 60 mM NaH2P04, 6 mM EDTA, 0.005% Triton X-100, pH 7.6)
at 22.degree. C. on a fluidics station (Affymetrix) for 10.times.2
cycles, and then washed with 0.1 MES at 45.degree. C. for 30 min.
The arrays are then stained with a streptavidin-phycoerythrin
conjugate (Molecular Probes), followed by 6.times.SSPE-T wash on
the fluidics station for 10.times.2 cycles again. To enhance the
signals, the arrays are further stained with Anti-streptavidin
antibody for 30 min followed by a 15 min staining with a
streptavidin-phycoerythrin conjugate again. After 6.times.SSPE-T
wash on the fluidics station for 10.times.2 cycles, the arrays are
scanned at a resolution of 3 microns using a modified confocal
scanner to determine raw fluorescence intensity values at each
position in the array, corresponding to gene expression levels for
the sequence at that array position.
[0343] The raw fluorescence intensity values are converted to gene
expression product levels, normalized via the RMA method, filtered
to remove data that may be considered suspect, and input to a
pre-classifier algorithm which corrects the gene expression product
levels for the cell-type composition of the biological sample. The
corrected gene expression product levels are input to a trained
algorithm for classifying the biological sample as benign,
malignant, or normal. The trained algorithm provides a record of
its output including a diagnosis, and a confidence level.
Example 10
Molecular Profiling of Thyroid Nodule
[0344] An individual notices a lump on his thyroid. The individual
consults his family physician. The family physician decides to
obtain a sample from the lump and subject it to molecular profiling
analysis. Said physician uses a kit from the molecular profiling
business to obtain the sample via fine needle aspiration, perform
an adequacy test, store the sample in a liquid based cytology
solution, and send it to the molecular profiling business. The
molecular profiling business divides the sample for cytological
analysis of one part and for the remainder of the sample extracts
mRNA from the sample, analyzes the quality and suitability of the
mRNA sample extracted, and analyses the expression levels and
alternative exon usage of a subset of the genes listed in FIG. 5.
In this case, the particular gene expression products profiled is
determined by the sample type, by the preliminary diagnosis of the
physician, and by the molecular profiling company.
[0345] The molecular profiling business analyses the data and
provides a resulting diagnosis to the individual's physician as
illustrated in FIG. 20. The results provide 1) a list of gene
expression products profiled, 2) the results of the profiling (e.g.
the expression level normalized to an internal standard such as
total mRNA or the expression of a well characterized gene product
such as tubulin, 3) the gene product expression level expected for
normal tissue of matching type, and 4) a diagnosis and recommended
treatment for Bob based on the gene product expression levels. The
molecular profiling business bills the individual's insurance
provider for products and services rendered.
Example 11
Molecular Profiling as an Adjunct to Cytological Examination
[0346] An individual notices a suspicious lump on her thyroid. The
individual consults her primary care physician who examines the
individual and refers her to an endocrinologist. The
endocrinologist obtains a sample via fine needle aspiration, and
sends the sample to a cytological testing laboratory. The
cytological testing laboratory performs routine cytological testing
on a portion of the fine needle aspirate, the results of which are
ambiguous (i.e. indeterminate). The cytological testing laboratory
suggests to the endocrinologist that the remaining sample may be
suitable for molecular profiling, and the endocrinologist
agrees.
[0347] The remaining sample is analyzed using the methods and
compositions herein. The results of the molecular profiling
analysis suggest a high probability of early stage follicular cell
carcinoma. The results further suggest that molecular profiling
analysis combined with patient data including patient age, and lump
or nodule size indicates thyroidectomy followed by radioactive
iodine ablation. The endocrinologist reviews the results and
prescribes the recommended therapy.
[0348] The cytological testing laboratory bills the endocrinologist
for routine cytological tests and for the molecular profiling. The
endocrinologist remits payment to the cytological testing
laboratory and bills the individual's insurance provider for all
products and services rendered. The cytological testing laboratory
passes on payment for molecular profiling to the molecular
profiling business and withholds a small differential.
Example 12
Molecular Profiling Performed by a Third Party
[0349] An individual complains to her physician about a suspicious
lump on her neck. The physician examines the individual, and
prescribes a molecular profiling test and a follow up examination
pending the results. The individual visits a clinical testing
laboratory also known as a CLIA lab. The CLIA lab is licensed to
perform molecular profiling of the current invention. The
individual provides a sample at the CLIA lab via fine needle
aspiration, and the sample is analyzed using the molecular
profiling methods and compositions herein. The results of the
molecular profiling are electronically communicated to the
individual's physician, and the individual is contacted to schedule
a follow up examination. The physician presents the results of the
molecular profiling to the individual and prescribes a therapy.
Example 13
Overlapping Genes Using Different Analysis Methods
[0350] The results described in Example 2 were obtained by
examining surgical thyroid nodule tissue samples and comparing gene
expression in malignant versus benign ("malignant vs. benign" data
set). This analysis identified 412 genes that are differentially
expressed (FDR p<0.05). In a previous study described in Example
1, using i) a different cohort of samples and ii) a different
analysis method, we describe 4918 genes that can distinguish
between malignant and benign thyroid nodules ("4918"). The
"malignant vs. benign" tissue discovery dataset shares 231/412
genes with the "4918" discovery dataset, while 181/412 genes have
been newly discovered.
[0351] A similar comparison between medullary thyroid cancer (MTC)
and the "Rest" of the thyroid subtypes using the tissue cohort
pointed to 668 significant genes that are differentially expressed
between these two groups (FIG. 10). When cross-checked against our
previous "4918" gene list, we note that 305/668 genes had been
previously described, while 363/668 genes have been newly
discovered.
[0352] We next combined the surgical tissue dataset with a fine
needle aspirate (FNA) dataset and once again compared malignant
versus benign using i) a "hot probes" and ii) a "Bayes" approach.
Each analysis identified 102 and 498 significant genes,
respectively (Tables 11 and 12).
[0353] Up until this point a total of 1343 significant genes were
identified. However, a subsequent subset analysis aimed at
identifying those genes that separate distinct pathology subtypes
from one another was also performed. This analysis used the
surgical tissue cohort and resulted in 250 significant genes (FIG.
13).
[0354] In sum, the five comparisons described here give rise to
1437 significant genes. Of these, 636/1437 genes are described for
the first time as distinguishing malignant versus benign thyroid
pathology. As of today, 568/636 have not yet been described in
published scientific literature or patent applications as
diagnostic markers of thyroid cancer.
Example 14
Clinical Thyroid FNA
[0355] Methods
[0356] Prospective clinical thyroid FNA samples were examined with
the Affymetrix Human Exon 1.0ST microarray in order to identify
genes that differ significantly in mRNA expression between benign
and malignant samples.
[0357] Affymetrix software was used to extract, normalize, and
summarize intensity data from roughly 6.5 million probes.
Approximately 280,000 core probe sets were subsequently used in
feature selection and classification. The models used were LIMMA
(for feature selection), random forest and SVM were used for
classification (Smyth 2004; Diaz-Uriarte and Alvarez de Andres
2006). Iterative rounds of training, classification, and
cross-validation were performed using random subsets of data. Top
features were identified in three separate analyses using the
classification engine described above.
[0358] While the annotation and mapping of genes to transcript
cluster indentifiers (TCID) is constantly evolving, the nucleotide
sequences in the probesets that make up a TCID do not change.
Furthermore, a number of significant TCIDs do not map any known
genes, yet these are equally important biomarkers in the
classification of thyroid malignancy. Results are described using
both the TCID and the genes currently mapped to each (Affymetrix
annotation file:
HuEx-1.sub.--0-st-v2.na29.hg18.transcript.csv).
[0359] Results
[0360] The study of differential gene expression in prospectively
collected, clinical thyroid FNA required a number of statistical
sub-analyses. These sub-analyses alone resulted in the discovery of
genes that are valuable in the classification of thyroid nodules of
unknown pathology. However, the joining of the datasets has
resulted in the novel characterization of thyroid gene panels,
which can correctly classify thyroid FNA with improved accuracy
over current cytopathology, and molecular profiling methods.
TABLE-US-00004 TABLE 3 Top Benign vs. Malignant Analysis. This
analysis resulted in 175 unique TCIDs, currently mapping to 198
genes. Gene Symbol (Affy FDR LIMMA TCID v.na29) p-value Fold Change
2884845 GABRB2 2.85E-35 3.22 2400177 CAMK2N1 8.23E-30 2.50 3638204
MFGE8 2.16E-29 1.75 3638204 QTRT1 2.16E-29 1.75 2708855 C11orf72
4.11E-27 2.27 2708855 LIPH 4.11E-27 2.27 3415744 IGFBP6 5.44E-27
1.81 3136178 PLAG1 1.64E-26 1.76 2657808 CLDN16 3.63E-26 3.01
3451375 PRICKLE1 3.63E-26 1.78 2442008 RXRG 7.62E-26 2.17 3329343
MDK 3.60E-24 1.34 3666366 CDH3 3.60E-24 1.25 3757108 KRT19 1.06E-23
1.44 3040518 MACC1 1.14E-23 1.73 3988596 ZCCHC12 2.14E-23 2.22
3416895 METTL7B 2.90E-23 1.33 2721959 ROS1 6.26E-23 3.05 2721959
SLC34A2 6.26E-23 3.05 3125116 DLC1 9.12E-23 0.82 2828441 PDLIM4
9.51E-23 0.81 2783596 PDE5A 1.60E-22 1.93 3645555 TNFRSF12A
1.71E-22 1.25 3973891 CXorf27 1.75E-22 1.38 3973891 SYTL5 1.75E-22
1.38 2827645 SLC27A6 2.02E-22 2.28 3020343 MET 2.02E-22 2.25
3452478 AMIGO2 2.03E-22 1.17 2451931 GOLT1A 2.15E-22 0.84 3679959
EMP2 3.81E-22 1.51 3417249 ERBB3 1.11E-21 1.05 3087167 TUSC3
1.16E-21 1.90 2924492 HEY2 1.38E-21 1.38 2685304 PROS1 1.48E-21
2.15 3335894 CST6 1.50E-21 2.50 3393720 MPZL2 1.52E-21 1.86 3907234
SDC4 1.60E-21 1.64 4012178 CITED1 4.03E-21 2.42 2994981 PRR15
5.89E-21 0.94 2973232 C6orf174 6.09E-21 1.07 2973232 KIAA0408
6.09E-21 1.07 2809245 ITGA2 6.13E-21 1.84 3067478 NRCAM 9.01E-21
1.70 3420316 HMGA2 1.13E-20 0.94 4018327 TRPC5 1.14E-20 1.78
3416921 RDH5 1.24E-20 0.55 2333318 PTPRF 1.42E-20 0.78 3336486
C11orf80 1.71E-20 0.58 3336486 RCE1 1.71E-20 0.58 3044072 NOD1
3.06E-20 1.01 3417809 NAB2 3.40E-20 0.57 2710599 CLDN1 4.47E-20
2.53 3343452 FZD4 4.93E-20 1.49 3343452 PRSS23 4.93E-20 1.49
2720584 SLIT2 6.84E-20 1.45 3389976 SLC35F2 1.16E-19 0.94 3587495
SCG5 1.45E-19 1.60 3744463 MYH10 1.58E-19 1.40 3987607 CCDC121
1.87E-19 1.56 3987607 ZCCHC16 1.87E-19 1.56 3984945 ARMCX3 3.69E-19
1.11 2558612 TGFA 9.18E-19 0.89 3522398 AIDA 1.02E-18 1.33 3522398
DOCK9 1.02E-18 1.33 2781736 CFI 1.04E-18 1.91 3338192 CCND1
1.09E-18 1.25 3338192 FLJ42258 1.09E-18 1.25 2414958 TACSTD2
1.12E-18 0.91 2991860 ITGB8 1.51E-18 1.30 2805078 CDH6 1.64E-18
1.58 3976341 TIMP1 1.98E-18 1.68 2562435 EDNRB 1.98E-18 1.61
2562435 SFTPB 1.98E-18 1.61 3726154 ITGA3 2.04E-18 1.17 2381249
C1orf115 4.38E-18 0.92 2356818 BCL9 6.05E-18 0.63 3451814 MAFG
7.13E-18 1.92 3451814 NELL2 7.13E-18 1.92 3445908 EPS8 7.19E-18
1.60 2451870 ETNK2 8.68E-18 1.00 3201345 LOC554202 1.08E-17 1.05
3581221 AHNAK2 1.14E-17 1.28 2966193 C6orf168 1.23E-17 0.85 2876608
CXCL14 1.85E-17 1.76 3129065 CLU 1.85E-17 1.37 3222170 TNC 1.94E-17
1.24 2438458 CRABP2 2.16E-17 1.24 2600689 EPHA4 2.17E-17 1.51
3763390 TMEM100 2.61E-17 1.34 2902958 C4A 3.56E-17 1.36 2902958 C4B
3.56E-17 1.36 2952834 KCNK5 6.07E-17 0.51 2452478 LEMD1 9.66E-17
1.27 3751002 RAB34 1.14E-16 0.83 3489138 CYSLTR2 1.72E-16 1.61
2417362 DIRAS3 1.72E-16 1.15 2370123 XPR1 1.81E-16 0.89 2680046
ADAMTS9 1.83E-16 1.40 3494629 SCEL 2.04E-16 1.61 3040967 RAPGEF5
2.04E-16 0.92 3554452 KIAA0284 2.33E-16 0.59 4020655 ODZ1 2.44E-16
1.97 2400518 ECE1 3.31E-16 0.98 2598261 FN1 3.58E-16 2.41 3187686
GSN 4.03E-16 0.78 2742224 SPRY1 3.51E-15 1.18 3628832 DAPK2
4.59E-15 1.17 3408831 SSPN 4.69E-15 0.99 3925639 NRIP1 5.01E-15
1.02 3683377 GPRC5B 5.39E-15 1.10 2397025 DHRS3 5.83E-15 1.14
2816298 IQGAP2 6.56E-15 -1.04 3848039 C3 7.85E-15 1.62 3367673
MPPED2 7.93E-15 -1.71 2822215 PAM 8.70E-15 1.08 2567167 LONRF2
1.12E-14 1.40 2522094 SPATS2L 2.21E-14 0.96 3898355 FLRT3 2.70E-14
1.96 3717870 TMEM98 2.72E-14 1.51 3212008 FRMD3 3.50E-14 1.43
2597867 IKZF2 3.58E-14 0.91 3007960 CLDN4 6.44E-14 1.27 2468811
ASAP2 7.11E-14 0.89 3046197 ELMO1 8.04E-14 -1.10 3132616 ZMAT4
8.04E-14 -1.29 3181600 GALNT12 8.25E-14 0.74 3095313 C8orf4
8.38E-14 1.28 2525533 LOC648149 8.38E-14 1.01 2525533 MAP2 8.38E-14
1.01 3464860 DUSP6 9.39E-14 1.10 3464860 LOC100131490 9.39E-14 1.10
2751936 GALNT7 1.52E-13 0.93 2578790 LRP1B 1.65E-13 -1.33 2700365
TM4SF1 2.19E-13 1.60 2598828 IGFBP5 2.87E-13 1.67 3126191 PSD3
3.12E-13 1.34 3979101 FAAH2 3.88E-13 0.68 3577612 SERPINA1 3.99E-13
1.12 3577612 SERPINA2 3.99E-13 1.12 3622934 MYEF2 4.25E-13 0.92
3622934 SLC24A5 4.25E-13 0.92 2738664 SGMS2 4.47E-13 1.13 3692999
MT1G 4.65E-13 -2.43 2902844 C2 7.40E-13 1.36 2902844 CFB 7.40E-13
1.36 3662201 MT1F 8.84E-13 -1.87 3662201 MT1H 8.84E-13 -1.87
3662201 MT1P2 8.84E-13 -1.87 2617188 ITGA9 1.07E-12 1.05 3401704
CCND2 1.09E-12 0.86 2562529 ST3GAL5 1.34E-12 0.88 2371139 LAMC2
1.53E-12 0.99 2626802 PTPRG 1.83E-12 1.06 2834282 STK32A 2.53E-12
1.23 2526806 FN1 3.12E-12 1.84 3111561 MAPK6 3.66E-12 -2.04 3111561
PKHD1L1 3.66E-12 -2.04 3238962 KIAA1217 7.24E-12 1.21 3238962 PRINS
7.24E-12 1.21 3110608 TM7SF4 7.72E-12 1.92 2466554 TPO 1.14E-11
-1.78 3126368 PSD3 2.30E-11 1.39 3558418 STXBP6 3.35E-11 0.94
2980449 IPCEF1 3.42E-11 -1.05 3907190 SLPI 4.25E-11 1.61 2955932
GPR110 5.17E-11 1.29 2976360 PERP 7.31E-11 1.31 2686023 DCBLD2
8.03E-11 0.98 2915828 NT5E 9.40E-11 1.19 3219621 CTNNAL1 1.17E-10
1.01 3971451 PHEX 1.39E-10 1.53 3417583 RBMS2 1.39E-10 1.09 2424102
CNN3 1.58E-10 1.07 3369931 RAG2 2.12E-10 -1.41 2730746 SLC4A4
2.24E-10 -1.21 3010503 CD36 2.91E-10 -1.42 3446137 LMO3 3.09E-10
1.44 3933536 TFF3 3.09E-10 -1.10 4021777 IGSF1 3.11E-10 1.55
3467949 SLC5A8 4.08E-10 -1.34 3288518 C10orf72 4.26E-10 1.18
2336891 DIO1 4.31E-10 -1.73 2498274 C2orf40 4.39E-10 1.71 2740067
ANK2 5.52E-10 -0.90 2924330 TPD52L1 6.04E-10 1.09 2427469 SLC16A4
6.71E-10 1.37 2727587 KIT 1.23E-09 -1.24 3464417 MGAT4C 1.45E-09
1.26 2331558 BMP8A 3.61E-09 -1.55 2711205 ATP13A4 6.51E-09 1.15
3142381 FABP4 7.25E-09 -1.59 3743551 CLDN7 8.01E-09 1.13 3662150
MT1M 8.06E-09 -1.47 3662150 MT1P3 8.06E-09 -1.47 3166644 TMEM215
9.05E-09 1.51 3087659 SLC7A2 1.32E-08 1.28 3321055 TEAD1 1.37E-07
1.10 3059667 SEMA3D 1.43E-07 -1.83
TABLE-US-00005 TABLE 4 Top Subtype Analysis This analysis resulted
in 599 unique TCIDs, currently mapping to 681 genes. Gene Symbol
TCID (Affy vna29) Subtype 1 Subtype 2 Subtype 3 Subtype 4 3153400
3153400 NHP_PTC 3749600 3749600 NHP_PTC 3726691 ABCC3 FA_FVPTC
3368940 ABTB2 NHP_PTC 3279058 ACBD7 NHP_PTC 2796553 ACSL1 NHP_PTC
3299504 ACTA2 NHP_PTC 3927480 ADAMTS5 NHP_PTC 2680046 ADAMTS9
NHP_PTC FA_FVPTC NHP_FVPTC LCT_REST 3252170 ADK NHP_PTC 3039791
AGR2 NHP_PTC 3581221 AHNAK2 NHP_PTC 2991233 AHR NHP_PTC 3522398
AIDA NHP_PTC NHP_FVPTC 3226138 AK1 NHP_PTC 3233049 AKR1C3 NHP_FVPTC
4009849 ALAS2 NHP_PTC 3611625 ALDH1A3 NHP_PTC 3169331 ALDH1B1
FA_FVPTC 3571727 ALDH6A1 FA_FVPTC 3452478 AMIGO2 NHP_PTC 4018454
AMOT NHP_PTC 2740067 ANK2 NHP_PTC NHP_FVPTC 3323748 ANO5 FA_FVPTC
NHP_FVPTC 3174816 ANXA1 NHP_PTC 2732844 ANXA3 NHP_PTC 2881747 ANXA6
NHP_FVPTC 3046062 AOAH NHP_PTC 2455418 AP3S1 NHP_PTC 4002809 APOO
FA_FVPTC NHP_FVPTC 3595594 AQP9 NHP_PTC 2734421 ARHGAP24 NHP_PTC
2632453 ARL13B NHP_PTC 2931391 ARL4A NHP_PTC 3984945 ARMCX3 NHP_PTC
4015838 ARMCX6 NHP_PTC 3321150 ARNTL NHP_PTC 3768474 ARSG NHP_FVPTC
2468811 ASAP2 NHP_PTC 2526759 ATIC NHP_PTC 2711225 ATP13A4 NHP_PTC
2711205 ATP13A4 NHP_PTC 3105749 ATP6V0D2 NHP_FVPTC 3824596 B3GNT3
NHP_PTC 2356818 BCL9 NHP_PTC 2608725 BHLHE40 NHP_PTC 3448088
BHLHE41 NHP_PTC 3772187 BIRC5 LCT_REST 2331558 BMP8A NHP_PTC
NHP_FVPTC 3926080 BTG3 NHP_PTC 3288518 C10orf72 FA_FVPTC NHP_FVPTC
2708855 C11orf72 NHP_PTC FA_FVPTC NHP_FVPTC 3327166 C11orf74
FA_FVPTC NHP_FVPTC 3336486 C11orf80 NHP_PTC 3473331 C12orf49
NHP_PTC 3571727 C14orf45 FA_FVPTC 3649714 C16orf45 NHP_PTC 3832280
C19orf33 NHP_PTC 2381249 C1orf115 NHP_PTC 2453065 C1orf116 NHP_PTC
2902844 C2 NHP_PTC 3963676 C22orf9 NHP_FVPTC 2498274 C2orf40
NHP_PTC FA_FVPTC NHP_FVPTC 3848039 C3 NHP_PTC 2902958 C4A NHP_PTC
FA_FVPTC NHP_FVPTC 2902958 C4B NHP_PTC FA_FVPTC NHP_FVPTC 2766492
C4orf34 NHP_PTC 2730303 C4orf7 LCT_REST 2855578 C5orf28 FA_FVPTC
2966193 C6orf168 NHP_PTC 2973232 C6orf174 NHP_PTC FA_FVPTC 3060450
C7orf62 NHP_PTC 3095313 C8orf4 NHP_PTC 3086809 C8orf79 FA_FVPTC
3867264 CA11 NHP_PTC 3392332 CADM1 NHP_PTC 2400177 CAMK2N1 NHP_PTC
FA_FVPTC NHP_FVPTC 3420713 CAND1 NHP_PTC 3020302 CAV1 NHP_PTC
3020273 CAV2 NHP_PTC 3987607 CCDC121 NHP_PTC FA_FVPTC NHP_FVPTC
2582701 CCDC148 NHP_PTC 2688813 CCDC80 NHP_PTC 3718204 CCL13
NHP_PTC 3204285 CCL19 LCT_REST 3338192 CCND1 NHP_PTC FA_FVPTC
NHP_FVPTC 3380065 CCND1 NHP_FVPTC 3401704 CCND2 NHP_PTC NHP_FVPTC
3316344 CD151 NHP_PTC 2860178 CD180 LCT_REST 2636125 CD200 NHP_PTC
3010503 CD36 NHP_PTC NHP_FVPTC 3834502 CD79A LCT_REST 2671728 CDCP1
NHP_PTC 3694657 CDH11 NHP_PTC 3666366 CDH3 NHP_PTC FA_FVPTC 2805078
CDH6 NHP_PTC 3417146 CDK2 NHP_PTC 2773719 CDKL2 NHP_PTC 2871896
CDO1 NHP_PTC 4024373 CDR1 NHP_PTC 2902844 CFB NHP_PTC 2373336 CFH
NHP_PTC 2373336 CFHR1 NHP_PTC 2781736 CFI NHP_PTC 3920003 CHAF1B
NHP_PTC 3442054 CHD4 NHP_PTC 4012178 CITED1 NHP_PTC FA_FVPTC
NHP_FVPTC 3178583 CKS2 NHP_PTC 3862108 CLC NHP_PTC 2710599 CLDN1
NHP_PTC NHP_FVPTC 3497195 CLDN10 NHP_PTC 2657808 CLDN16 NHP_PTC
FA_FVPTC NHP_FVPTC 3007960 CLDN4 NHP_PTC NHP_FVPTC 3743551 CLDN7
NHP_PTC 3443183 CLEC4E NHP_PTC 3129065 CLU NHP_PTC FA_FVPTC 2424102
CNN3 NHP_PTC 3762198 COL1A1 NHP_PTC 3761054 COPZ2 FA_FVPTC
NHP_FVPTC 3106559 CP NHP_PTC 3105904 CPNE3 FA_FVPTC 2377283 CR2
LCT_REST 3603295 CRABP1 NHP_PTC 2438458 CRABP2 NHP_PTC FA_FVPTC
2406783 CSF3R NHP_PTC 3126504 CSGALNACT1 FA_FVPTC 3335894 CST6
NHP_PTC FA_FVPTC NHP_FVPTC 3219621 CTNNAL1 NHP_PTC 2618940 CTNNB1
NHP_PTC 3634811 CTSH NHP_PTC 3338552 CTTN NHP_PTC 2773434 CXCL1
NHP_PTC 2732508 CXCL13 LCT_REST 2876608 CXCL14 NHP_PTC 3863640
CXCL17 NHP_PTC 2773434 CXCL2 NHP_PTC 2773434 CXCL3 NHP_PTC 4024420
CXorf18 NHP_PTC 3973891 CXorf27 NHP_PTC 3910429 CYP24A1 NHP_FVPTC
2528093 CYP27A1 NHP_FVPTC 3489138 CYSLTR2 NHP_PTC FA_FVPTC
NHP_FVPTC 3628832 DAPK2 NHP_PTC FA_FVPTC NHP_FVPTC 2686023 DCBLD2
NHP_PTC 3683845 DCUN1D3 NHP_PTC 2420832 DDAH1 NHP_PTC 3329649 DDB2
NHP_PTC 3754736 DDX52 NHP_PTC 3487095 DGKH NHP_PTC 3074912 DGKI
NHP_PTC 3558118 DHRS1 NHP_PTC 2397025 DHRS3 NHP_PTC 2336891 DIO1
NHP_PTC FA_FVPTC NHP_FVPTC 2417362 DIRAS3 NHP_PTC NHP_FVPTC 3125116
DLC1 NHP_PTC FA_FVPTC 3522398 DOCK9 NHP_PTC NHP_FVPTC 3913483 DPH3B
FA_FVPTC 2584018 DPP4 NHP_PTC 2880292 DPYSL3 NHP_PTC 3783529 DSG2
NHP_PTC 2893794 DSP NHP_PTC 2958325 DST NHP_PTC 3622176 DUOX1
NHP_FVPTC 3622176 DUOX2 NHP_FVPTC 3622239 DUOXA1 NHP_FVPTC 3622239
DUOXA2 NHP_FVPTC 3129731 DUSP4 NHP_PTC 3263743 DUSP5 NHP_PTC
3464860 DUSP6 NHP_PTC 3497195 DZIP1 NHP_PTC 2400518 ECE1 NHP_PTC
FA_FVPTC NHP_FVPTC 2562435 EDNRB NHP_PTC 3002640 EGFR NHP_PTC
2484970 EHBP1 NHP_PTC 3837431 EHD2 NHP_PTC 3326461 EHF NHP_PTC
3544387 EIF2B2 FA_FVPTC 3427098 ELK3 NHP_PTC 3046197 ELMO1 NHP_PTC
3679959 EMP2 NHP_PTC FA_FVPTC NHP_FVPTC 3852832 EMR3 NHP_PTC
2458338 ENAH NHP_PTC 3345427 ENDOD1 NHP_PTC 2327677 EPB41 NHP_PTC
2600689 EPHA4 NHP_PTC 2346625 EPHX4 NHP_PTC 3772187 EPR1 LCT_REST
3445908 EPS8 NHP_PTC 3720402 ERBB2 FA_FVPTC 3417249 ERBB3 NHP_PTC
3683845 ERI2 NHP_PTC 2462329 ERO1LB FA_FVPTC 3445768 ERP27 NHP_PTC
2451870 ETNK2 NHP_PTC 3039177 ETV1 NHP_PTC 2709132 ETV5 NHP_PTC
2863363 F2RL2 NHP_PTC 3979101 FAAH2 NHP_PTC 3142381 FABP4 NHP_PTC
NHP_FVPTC 3331926 FAM111A NHP_PTC 3331903 FAM111B NHP_PTC 3104323
FAM164A NHP_PTC 2560625 FAM176A NHP_PTC 3768535 FAM20A NHP_PTC
3143330 FAM82B FA_FVPTC 3152558 FAM84B NHP_PTC 2396750 FBXO2
NHP_PTC 3473480 FBXO21 NHP_PTC 3229338 FCN1 NHP_PTC 3229338 FCN2
NHP_PTC 2742109 FGF2 NHP_PTC 3413950 FGFR1OP2 NHP_PTC 3324447 FIBIN
FA_FVPTC 2738244 FLJ20184 NHP_PTC 3346147 FLJ32810 NHP_PTC 3338192
FLJ42258 NHP_PTC FA_FVPTC NHP_FVPTC 3380065 FLJ42258 NHP_FVPTC
3898355 FLRT3 NHP_PTC 2526806 FN1 NHP_PTC 2598261 FN1 NHP_PTC
FA_FVPTC 3869237 FPR1 NHP_PTC 3839910 FPR2 NHP_PTC 3212008 FRMD3
NHP_PTC FA_FVPTC NHP_FVPTC 3393479 FXYD6 NHP_FVPTC 3343452 FZD4
NHP_PTC 3110272 FZD6 NHP_PTC 2523045 FZD7 NHP_PTC 3217242 GABBR2
NHP_PTC 2884845 GABRB2 NHP_PTC FA_FVPTC NHP_FVPTC 2341083 GADD45A
NHP_PTC 2401581 GALE NHP_PTC 3181600 GALNT12 NHP_PTC 2585129 GALNT3
NHP_PTC 2751936 GALNT7 NHP_PTC 2684187 GBE1 NHP_FVPTC 2421843 GBP1
NHP_PTC 2421843 GBP3 NHP_PTC 3044129 GGCT NHP_PTC 4015763 GLA
NHP_FVPTC 3593931 GLDN NHP_PTC 2417272 GNG12 NHP_PTC
2451931 GOLT1A NHP_PTC 2955932 GPR110 NHP_PTC 2955999 GPR110
NHP_PTC 2819779 GPR98 NHP_PTC NHP_FVPTC 3683377 GPRC5B NHP_PTC
2827057 GRAMD3 NHP_PTC 3187686 GSN NHP_PTC 2787958 GYPB NHP_PTC
2504328 GYPC NHP_PTC 2787958 GYPE NHP_PTC 2809793 GZMK LCT_REST
3217077 HEMGN NHP_PTC 2924492 HEY2 NHP_PTC FA_FVPTC NHP_FVPTC
2946194 HIST1H1A NHP_PTC 2946215 HIST1H3B LCT_REST 2947081 HIST1H4L
LCT_REST 2950125 HLA-DQB2 NHP_PTC 3420316 HMGA2 NHP_PTC 3830065 HPN
NHP_PTC 2658275 HRASLS FA_FVPTC 3508330 HSPH1 NHP_PTC 3820443 ICAM1
NHP_PTC 2401493 ID3 FA_FVPTC 2708922 IGF2BP2 FA_FVPTC 2598828
IGFBP5 NHP_PTC 3415744 IGFBP6 NHP_PTC FA_FVPTC NHP_FVPTC 4021777
IGSF1 NHP_PTC FA_FVPTC 2597867 IKZF2 FA_FVPTC NHP_FVPTC 3755862
IKZF3 FA_FVPTC 2497082 IL1RL1 NHP_PTC 3275729 IL2RA NHP_FVPTC
2731332 IL8 NHP_FVPTC 2599303 IL8RA NHP_PTC 2599303 IL8RB NHP_PTC
2527580 IL8RB NHP_PTC 2599303 IL8RBP NHP_PTC 2527580 IL8RBP NHP_PTC
2673873 IMPDH2 NHP_PTC 3267382 INPP5F NHP_PTC 2980449 IPCEF1
NHP_PTC 2816298 IQGAP2 NHP_PTC 2809245 ITGA2 NHP_PTC 3726154 ITGA3
NHP_PTC 2617188 ITGA9 NHP_PTC FA_FVPTC 3852832 ITGB1 NHP_PTC
2583465 ITGB6 NHP_PTC 2991860 ITGB8 NHP_PTC 4013549 ITM2A LCT_REST
2608469 ITPR1 NHP_PTC 3556990 JUB NHP_PTC 3998766 KAL1 NHP_PTC
2628260 KBTBD8 LCT_REST 2952834 KCNK5 NHP_PTC 3154002 KCNQ3 NHP_PTC
3383130 KCTD14 NHP_PTC 2827525 KDELC1 NHP_PTC 3945314 KDELR3
NHP_PTC 2959039 KHDRBS2 NHP_PTC 3554452 KIAA0284 NHP_PTC NHP_FVPTC
2973232 KIAA0408 NHP_PTC FA_FVPTC 3238962 KIAA1217 NHP_PTC
NHP_FVPTC 3529951 KIAA1305 FA_FVPTC 2727587 KIT NHP_PTC FA_FVPTC
NHP_FVPTC 3978943 KLF8 NHP_PTC 2708066 KLHL6 LCT_REST 3868828 KLK10
NHP_PTC 3868783 KLK7 NHP_PTC 3415576 KRT18 NHP_PTC 3757108 KRT19
NHP_PTC FA_FVPTC 2453793 LAMB3 NHP_PTC 2371065 LAMC1 NHP_PTC
2371139 LAMC2 NHP_PTC 2962026 LCA5 NHP_PTC 3653619 LCMT1 NHP_PTC
3190190 LCN2 NHP_PTC 4024420 LDOC1 NHP_PTC 2452478 LEMD1 NHP_PTC
2854092 LIFR FA_FVPTC 3841545 LILRA1 NHP_PTC 3841545 LILRB1 NHP_PTC
3454331 LIMA1 NHP_PTC 3202528 LINGO2 NHP_PTC 2708855 LIPH NHP_PTC
FA_FVPTC NHP_FVPTC 3446137 LMO3 NHP_PTC FA_FVPTC NHP_FVPTC 2345286
LMO4 NHP_PTC 3028011 LOC100124692 NHP_PTC 3442054 LOC100127974
NHP_PTC 3765689 LOC100129112 NHP_PTC 3759587 LOC100129115 NHP_PTC
2601414 LOC100129171 NHP_PTC 2577482 LOC100129961 NHP_PTC 2504328
LOC100130248 NHP_PTC 3110272 LOC100131102 NHP_PTC 3464860
LOC100131490 NHP_PTC 2364677 LOC100131938 NHP_PTC 3922793
LOC100132338 NHP_PTC 3392332 LOC100132764 NHP_PTC 3487095 LOC283508
NHP_PTC 3724698 LOC440434 NHP_PTC 3201345 LOC554202 NHP_PTC 2455418
LOC643454 NHP_PTC 2525533 LOC648149 NHP_PTC 4015838 LOC653354
NHP_PTC 3724698 LOC653498 NHP_PTC 2936857 LOC730031 NHP_PTC 2567167
LONRF2 NHP_PTC LCT_REST 2872848 LOX NHP_PTC 3220384 LPAR1 NHP_FVPTC
3442137 LPAR5 NHP_PTC 3088486 LPL NHP_PTC 2578790 LRP1B NHP_PTC
NHP_FVPTC 3106559 LRRC69 NHP_PTC 2608309 LRRN1 NHP_PTC 3465248 LUM
NHP_PTC 3683845 LYRM1 NHP_PTC 3040518 MACC1 NHP_PTC FA_FVPTC
3451814 MAFG NHP_PTC FA_FVPTC 3994710 MAMLD1 NHP_PTC 2525533 MAP2
NHP_PTC 3111561 MAPK6 NHP_PTC NHP_FVPTC 3108526 MATN2 FA_FVPTC
2539607 MBOAT2 NHP_PTC 3097152 MCM4 NHP_PTC 3063685 MCM7 NHP_PTC
3329343 MDK NHP_PTC FA_FVPTC 2962820 ME1 NHP_FVPTC 3765689 MED13
NHP_PTC 3020343 MET NHP_PTC FA_FVPTC NHP_FVPTC 3416895 METTL7B
NHP_PTC FA_FVPTC NHP_FVPTC 3808096 MEX3C NHP_PTC 3638204 MFGE8
NHP_PTC FA_FVPTC NHP_FVPTC 3028011 MGAM NHP_PTC 2890859 MGAT1
NHP_FVPTC 3464417 MGAT4C NHP_PTC 2658275 MGC2889 FA_FVPTC 3406589
MGST1 NHP_PTC 3707759 MIS12 FA_FVPTC 2936857 MLLT4 NHP_PTC 3143660
MMP16 NHP_PTC 3143643 MMP16 NHP_PTC 2362333 MNDA NHP_PTC 4017212
MORC4 NHP_PTC 3367673 MPPED2 NHP_PTC 3393720 MPZL2 NHP_PTC FA_FVPTC
NHP_FVPTC 2955025 MRPL14 NHP_PTC 3662201 MT1F NHP_PTC FA_FVPTC
3692999 MT1G NHP_PTC NHP_FVPTC 3662201 MT1H NHP_PTC FA_FVPTC
3662150 MT1M NHP_PTC 3662201 MT1P2 NHP_PTC FA_FVPTC 3662150 MT1P3
NHP_PTC 2931391 MTHFD1L NHP_PTC 2437118 MUC1 NHP_PTC 3366903 MUC15
NHP_PTC 3655723 MVP NHP_PTC 3997825 MXRA5 NHP_PTC 3622934 MYEF2
NHP_PTC 3744463 MYH10 NHP_PTC 2520429 MYO1B NHP_PTC 3752709 MYO1D
NHP_PTC 3624607 MYO5A NHP_FVPTC 2914070 MYO6 NHP_PTC 3417809 NAB2
NHP_PTC 3695268 NAE1 NHP_PTC 3074912 NAG20 NHP_PTC 3323052 NAV2
FA_FVPTC NHP_FVPTC 3349293 NCAM1 FA_FVPTC 2590736 NCKAP1 NHP_PTC
3495076 NDFIP2 NHP_PTC 3789947 NEDD4L NHP_PTC 3451814 NELL2 NHP_PTC
FA_FVPTC 2343231 NEXN NHP_PTC 3456666 NFE2 NHP_PTC 3199207 NFIB
NHP_PTC 2325410 NIPAL3 NHP_PTC 3182957 NIPSNAP3A FA_FVPTC 3182957
NIPSNAP3B FA_FVPTC 3044072 NOD1 NHP_PTC FA_FVPTC 3571904 NPC2
NHP_PTC 3724698 NPEPPS NHP_PTC 2370926 NPL NHP_FVPTC 2792127 NPY1R
NHP_PTC 3067478 NRCAM NHP_PTC NHP_FVPTC 3925639 NRIP1 NHP_PTC
2524301 NRP2 NHP_PTC 2915828 NT5E NHP_PTC 3143330 NTAN1 FA_FVPTC
3322251 NUCB2 FA_FVPTC NHP_FVPTC 2742109 NUDT6 NHP_PTC 3654699
NUPR1 FA_FVPTC 2768654 OCIAD2 NHP_PTC 2375338 OCR1 NHP_PTC 4020655
ODZ1 NHP_PTC FA_FVPTC NHP_FVPTC 3380065 ORAOV1 NHP_FVPTC 3801621
OSBPL1A NHP_FVPTC 3555461 OSGEP NHP_PTC 2807359 OSMR NHP_PTC
2701071 P2RY13 NHP_PTC 2875193 P4HA2 NHP_PTC 2822215 PAM NHP_PTC
3256590 PAPSS2 NHP_FVPTC 3505781 PARP4 NHP_PTC 3320865 PARVA
NHP_PTC 2364677 PBX1 NHP_PTC 3134922 PCMTD1 FA_FVPTC 2783596 PDE5A
NHP_PTC NHP_FVPTC 3922793 PDE9A NHP_PTC 3087703 PDGFRL NHP_PTC
3301218 PDLIM1 NHP_PTC 2828441 PDLIM4 NHP_PTC 3411810 PDZRN4
NHP_PTC 3013255 PEG10 NHP_PTC 2976360 PERP NHP_PTC 3971451 PHEX
NHP_PTC 3975893 PHF16 NHP_PTC 2635906 PHLDB2 NHP_PTC 3127385 PHYHIP
NHP_PTC 3811086 PIGN FA_FVPTC 3111561 PKHD1L1 NHP_PTC NHP_FVPTC
2511820 PKP4 NHP_PTC 3376529 PLA2G16 NHP_PTC 2955827 PLA2G7
NHP_FVPTC 2583374 PLA2R1 NHP_PTC 3136178 PLAG1 NHP_PTC FA_FVPTC
NHP_FVPTC 3252036 PLAU NHP_PTC 3759587 PLCD3 NHP_PTC 2521574 PLCL1
NHP_FVPTC 3867458 PLEKHA4 NHP_PTC 3407096 PLEKHA5 NHP_PTC 2858023
PLK2 NHP_PTC 3987996 PLS3 NHP_PTC 3911217 PMEPA1 NHP_PTC 3061997
PON2 NHP_PTC 2763550 PPARGC1A NHP_PTC 2773358 PPBP NHP_PTC 3678462
PPL NHP_PTC 2931090 PPP1R14C NHP_PTC 3384270 PRCP NHP_FVPTC 3451375
PRICKLE1 NHP_PTC 3238962 PRINS NHP_PTC NHP_FVPTC 2682271 PROK2
NHP_PTC 2685304 PROS1 NHP_PTC 2994981 PRR15 NHP_PTC 3973692 PRRG1
NHP_PTC 3343452 PRSS23 NHP_PTC 3175971 PSAT1 FA_FVPTC 3126368 PSD3
NHP_PTC 3126191 PSD3 NHP_PTC 2455418 PTPN14 NHP_PTC 2333318 PTPRF
NHP_PTC 2626802 PTPRG NHP_PTC
2973376 PTPRK NHP_PTC 3757917 PTRF NHP_PTC 3134922 PXDNL FA_FVPTC
3638204 QTRT1 NHP_PTC FA_FVPTC NHP_FVPTC 2361257 RAB25 NHP_PTC
3625271 RAB27A NHP_PTC 2929699 RAB32 NHP_FVPTC 3751002 RAB34
NHP_PTC 3183757 RAD23B NHP_PTC 3369931 RAG2 NHP_PTC FA_FVPTC
4001223 RAI2 NHP_PTC 3040967 RAPGEF5 NHP_PTC 3456081 RARG NHP_PTC
2819044 RASA1 NHP_PTC 3944210 RASD2 NHP_PTC 4000944 RBBP7 NHP_PTC
3781429 RBBP8 NHP_PTC 3417583 RBMS2 NHP_PTC 3336486 RCE1 NHP_PTC
3416921 RDH5 NHP_PTC 2779335 RG9MTD2 FA_FVPTC 2372812 RGS13
LCT_REST 2372719 RGS18 NHP_PTC 2372858 RGS2 NHP_PTC 2384401 RHOU
NHP_PTC 2580802 RND3 NHP_PTC 2721959 ROS1 NHP_PTC FA_FVPTC
NHP_FVPTC 2709606 RPL39L NHP_PTC 3804143 RPRD1A NHP_PTC 3867965
RRAS NHP_PTC 2469252 RRM2 LCT_REST 2442008 RXRG NHP_PTC FA_FVPTC
NHP_FVPTC 2435981 S100A12 NHP_PTC 4045665 S100A14 NHP_PTC 4045643
S100A16 NHP_PTC 2435989 S100A8 NHP_PTC 2359664 S100A9 NHP_PTC
3691326 SALL1 NHP_PTC NHP_FVPTC 3564027 SAV1 NHP_PTC 2750594 SC4MOL
NHP_PTC 3091475 SCARA3 NHP_PTC 3442054 SCARNA11 NHP_PTC 3494629
SCEL NHP_PTC 3587495 SCG5 NHP_PTC NHP_FVPTC 3441885 SCNN1A NHP_PTC
3043895 SCRN1 NHP_PTC 3907234 SDC4 NHP_PTC FA_FVPTC NHP_FVPTC
3779756 SEH1L NHP_PTC 2443450 SELL NHP_PTC 3058759 SEMA3C NHP_FVPTC
3059667 SEMA3D NHP_PTC 2732273 SEPT11 NHP_PTC 2328273 SERINC2
NHP_PTC 3577612 SERPINA1 NHP_PTC FA_FVPTC 3577612 SERPINA2 NHP_PTC
FA_FVPTC 2601414 SERPINE2 NHP_PTC 3331355 SERPING1 NHP_PTC 2326774
SFN NHP_PTC 2562435 SFTPB NHP_PTC 2768981 SGCB NHP_PTC 3061805 SGCE
NHP_PTC 2648535 SGEF NHP_PTC 2738664 SGMS2 NHP_PTC 3088213 SH2D4A
NHP_PTC 3304970 SH3PXD2A NHP_PTC 3894727 SIRPA NHP_PTC 3894727
SIRPB1 NHP_PTC 3154263 SLA NHP_PTC 2827525 SLC12A2 NHP_PTC 2427469
SLC16A4 NHP_PTC 3768412 SLC16A6 NHP_FVPTC 2960955 SLC17A5 NHP_PTC
3622934 SLC24A5 NHP_PTC 3018605 SLC26A4 NHP_PTC 3106559 SLC26A7
NHP_PTC 3593575 SLC27A2 NHP_PTC 2827645 SLC27A6 NHP_PTC 2721959
SLC34A2 NHP_PTC FA_FVPTC NHP_FVPTC 3216276 SLC35D2 FA_FVPTC 3389976
SLC35F2 NHP_PTC 3804195 SLC39A6 NHP_PTC 2730746 SLC4A4 NHP_PTC
3467949 SLC5A8 NHP_PTC 2786322 SLC7A11 FA_FVPTC NHP_FVPTC 3087659
SLC7A2 NHP_PTC 2720584 SLIT2 NHP_PTC 3907190 SLPI NHP_PTC 3509842
SMAD9 FA_FVPTC 2937144 SMOC2 NHP_PTC 3766960 SMURF2 NHP_PTC 2777714
SNCA NHP_PTC 3597857 SNX1 NHP_PTC 3597914 SNX22 NHP_PTC 2348437
SNX7 NHP_PTC 2369557 SOAT1 NHP_FVPTC 2797202 SORBS2 NHP_PTC 3413950
SPATS2 NHP_PTC 2522094 SPATS2L NHP_PTC 2585933 SPC25 LCT_REST
3590164 SPINT1 NHP_PTC 2556752 SPRED2 NHP_PTC 2742224 SPRY1 NHP_PTC
FA_FVPTC 3519309 SPRY2 NHP_PTC 3677969 SRL NHP_PTC 3408831 SSPN
NHP_PTC 2562529 ST3GAL5 NHP_PTC 3011861 STEAP2 FA_FVPTC 2834282
STK32A NHP_PTC NHP_FVPTC 3558418 STXBP6 NHP_FVPTC 3102372 SULF1
NHP_PTC 2979871 SYNE1 NHP_PTC 2378256 SYT14 NHP_PTC 3973891 SYTL5
NHP_PTC 2414958 TACSTD2 NHP_PTC 3898126 TASP1 FA_FVPTC 3724698
TBC1D3F NHP_PTC 3264621 TCF7L2 FA_FVPTC 3913483 TCFL5 FA_FVPTC
2435218 TDRKH NHP_PTC 3320944 TEAD1 NHP_PTC 3321055 TEAD1 NHP_PTC
2573570 TFCP2L1 NHP_PTC 3933536 TFF3 NHP_PTC 2591421 TFPI NHP_FVPTC
2558612 TGFA NHP_PTC 2380590 TGFB2 NHP_PTC 3181728 TGFBR1 NHP_PTC
3976341 TIMP1 NHP_PTC FA_FVPTC 2649113 TIPARP NHP_PTC 3615579 TJP1
NHP_PTC 3173880 TJP2 NHP_PTC 3751042 TLCD1 NHP_PTC 3969115 TLR8
NHP_PTC 2700365 TM4SF1 NHP_PTC 2647315 TM4SF4 NHP_PTC 3110608
TM7SF4 NHP_PTC 3763390 TMEM100 NHP_PTC 3412345 TMEM117 NHP_PTC
3346147 TMEM133 NHP_PTC 2577482 TMEM163 NHP_PTC 2815220 TMEM171
FA_FVPTC NHP_FVPTC 3166644 TMEM215 NHP_PTC NHP_FVPTC 3571904
TMEM90A NHP_PTC 3717870 TMEM98 NHP_PTC 3351200 TMPRSS4 NHP_PTC
3222170 TNC NHP_PTC 3150455 TNFRSF11B FA_FVPTC 3645555 TNFRSF12A
NHP_PTC FA_FVPTC NHP_FVPTC 3648391 TNFRSF17 LCT_REST 3222128
TNFSF15 NHP_PTC 3907111 TOMM34 NHP_PTC 3136888 TOX LCT_REST 2924330
TPD52L1 NHP_PTC 2466554 TPO NHP_PTC FA_FVPTC NHP_FVPTC 3818515
TRIP10 NHP_PTC 4018327 TRPC5 NHP_PTC FA_FVPTC NHP_FVPTC 3512294
TSC22D1 NHP_PTC 2991150 TSPAN13 NHP_PTC 4015397 TSPAN6 NHP_PTC
3891342 TUBB1 NHP_PTC 3779579 TUBB6 NHP_PTC 3401217 TULP3 NHP_PTC
3087167 TUSC3 NHP_PTC 3809324 TXNL1 FA_FVPTC 3429460 TXNRD1
NHP_FVPTC 3775842 TYMS NHP_PTC 2448971 UCHL5 NHP_FVPTC 2974592 VNN1
NHP_FVPTC 2974635 VNN2 NHP_PTC 2974610 VNN3 NHP_PTC 3203855 WDR40A
NHP_PTC 2489228 WDR54 NHP_PTC 3625052 WDR72 FA_FVPTC 3768474 WIPI1
NHP_FVPTC 2677356 WNT5A NHP_PTC 4015548 XKRX NHP_PTC 2370123 XPR1
NHP_PTC 3832280 YIF1B NHP_PTC 2413484 YIPF1 FA_FVPTC 4024373 YTHDC2
NHP_PTC 3989089 ZBTB33 NHP_PTC 3988596 ZCCHC12 NHP_PTC FA_FVPTC
NHP_FVPTC 3987607 ZCCHC16 NHP_PTC FA_FVPTC NHP_FVPTC 3569754
ZFP36L1 NHP_PTC 2706791 ZMAT3 NHP_PTC 3132616 ZMAT4 NHP_PTC
FA_FVPTC NHP_FVPTC 2331903 ZNF643 NHP_PTC 3011675 ZNF804B
NHP_PTC
TABLE-US-00006 TABLE 5 Trident Analysis This benign vs. malignant
analysis resulted in 210 unique TCIDs, currently mapping to 237
genes. These genes represent the union of three statistically
significant sub-analyses (Repeatable, Bayes, and Tissue) using a
single dataset. Gene Symbol TCID (Affy v.na29) Repeatable Bayes
Tissue DE P value 3393720 MPZL2 TRUE TRUE TRUE 1.49 1.87E-32
2400177 CAMK2N1 TRUE TRUE FALSE 1.67 2.27E-29 3067478 NRCAM TRUE
TRUE FALSE 1.42 2.53E-29 3445908 EPS8 TRUE TRUE TRUE 1.44 6.34E-29
3020343 MET TRUE TRUE FALSE 1.49 1.47E-27 4012178 CITED1 TRUE TRUE
FALSE 1.50 2.37E-27 2710599 CLDN1 TRUE TRUE FALSE 1.41 9.07E-27
3338192 CCND1 TRUE TRUE TRUE 1.36 2.63E-26 3338192 FLJ42258 TRUE
TRUE TRUE 1.36 2.63E-26 3126191 PSD3 TRUE TRUE TRUE 1.32 3.49E-25
2884845 GABRB2 TRUE TRUE FALSE 1.73 4.07E-25 3087167 TUSC3 TRUE
TRUE FALSE 1.49 6.22E-25 3907234 SDC4 TRUE TRUE FALSE 1.46 2.08E-24
2721959 ROS1 TRUE TRUE FALSE 1.48 2.82E-24 2721959 SLC34A2 TRUE
TRUE FALSE 1.48 2.82E-24 3679959 EMP2 FALSE TRUE FALSE 1.50
2.83E-24 2708855 C11orf72 TRUE TRUE TRUE 1.59 1.31E-23 2708855 LIPH
TRUE TRUE TRUE 1.59 1.31E-23 3416895 METTL7B TRUE TRUE FALSE 1.49
2.12E-23 3136178 PLAG1 FALSE TRUE TRUE 1.41 2.37E-23 2442008 RXRG
TRUE TRUE FALSE 1.60 3.50E-23 2657808 CLDN16 TRUE TRUE TRUE 1.51
2.63E-22 3984945 ARMCX3 TRUE TRUE FALSE 1.45 3.13E-22 2567167
LONRF2 TRUE TRUE TRUE 1.38 3.67E-22 2685304 PROS1 TRUE TRUE FALSE
1.46 3.81E-22 3744463 MYH10 TRUE TRUE FALSE 1.46 6.20E-22 3415744
IGFBP6 TRUE TRUE FALSE 1.56 9.91E-22 2834282 STK32A TRUE TRUE TRUE
1.27 1.03E-21 3554452 KIAA0284 TRUE TRUE FALSE 1.32 1.38E-21
3040518 MACC1 TRUE TRUE FALSE 1.47 1.42E-21 3587495 SCG5 TRUE TRUE
FALSE 1.34 1.74E-21 2686023 DCBLD2 TRUE TRUE FALSE 1.18 1.83E-21
3335894 CST6 FALSE TRUE FALSE 1.43 2.29E-21 2783596 PDE5A TRUE TRUE
TRUE 1.55 2.63E-21 3522398 AIDA TRUE TRUE FALSE 1.38 2.99E-21
3522398 DOCK9 TRUE TRUE FALSE 1.38 2.99E-21 3638204 MFGE8 TRUE TRUE
FALSE 1.51 5.35E-21 3638204 QTRT1 TRUE TRUE FALSE 1.51 5.35E-21
3323052 NAV2 TRUE TRUE FALSE 1.30 7.00E-21 2924492 HEY2 TRUE TRUE
FALSE 1.48 2.01E-20 3726154 ITGA3 TRUE TRUE FALSE 1.35 2.16E-20
2924330 TPD52L1 TRUE TRUE FALSE 1.17 2.21E-20 3988596 ZCCHC12 TRUE
TRUE FALSE 1.52 2.85E-20 3683377 GPRC5B TRUE TRUE FALSE 1.28
4.84E-20 3417249 ERBB3 FALSE TRUE FALSE 1.51 6.63E-20 2511820 PKP4
TRUE TRUE TRUE 1.22 7.51E-20 4020655 ODZ1 TRUE TRUE FALSE 1.34
8.32E-20 3628832 DAPK2 FALSE TRUE FALSE 1.34 1.20E-19 3007960 CLDN4
TRUE TRUE FALSE 1.20 1.42E-19 2598261 FN1 TRUE TRUE FALSE 1.31
3.25E-19 2936857 LOC730031 TRUE TRUE TRUE 1.12 5.16E-19 2936857
MLLT4 TRUE TRUE TRUE 1.12 5.16E-19 3666366 CDH3 TRUE TRUE TRUE 1.48
6.10E-19 3757108 KRT19 TRUE TRUE FALSE 1.41 6.20E-19 3451375
PRICKLE1 FALSE TRUE TRUE 1.42 8.79E-19 3338552 CTTN TRUE TRUE TRUE
1.10 9.53E-19 2680046 ADAMTS9 TRUE TRUE FALSE 1.40 1.06E-18 3867458
PLEKHA4 TRUE TRUE FALSE 1.35 1.50E-18 3494629 SCEL TRUE TRUE FALSE
1.39 1.57E-18 3978943 KLF8 TRUE TRUE FALSE 1.35 3.66E-18 2397025
DHRS3 TRUE TRUE FALSE 1.22 3.89E-18 3420316 HMGA2 TRUE TRUE TRUE
1.48 4.63E-18 3126368 PSD3 TRUE TRUE FALSE 1.19 5.77E-18 2809245
ITGA2 TRUE TRUE FALSE 1.45 6.16E-18 2526806 FN1 TRUE TRUE TRUE 1.19
7.55E-18 2827645 SLC27A6 FALSE TRUE FALSE 1.49 8.33E-18 3217361
ANKS6 TRUE TRUE FALSE 1.19 8.37E-18 3743551 CLDN7 TRUE TRUE FALSE
1.07 1.80E-17 3571904 NPC2 FALSE TRUE FALSE 0.99 2.53E-17 3571904
TMEM90A FALSE TRUE FALSE 0.99 2.53E-17 2558612 TGFA TRUE TRUE FALSE
1.35 2.71E-17 3987607 CCDC121 TRUE TRUE FALSE 1.46 3.28E-17 3987607
ZCCHC16 TRUE TRUE FALSE 1.46 3.28E-17 3088213 SH2D4A TRUE TRUE
FALSE 1.18 5.07E-17 3751002 RAB34 TRUE TRUE FALSE 1.19 5.77E-17
3973891 CXorf27 TRUE TRUE FALSE 1.52 6.03E-17 3973891 SYTL5 TRUE
TRUE FALSE 1.52 6.03E-17 3044072 NOD1 TRUE TRUE TRUE 1.45 6.85E-17
2370123 XPR1 TRUE TRUE FALSE 1.26 7.13E-17 3174816 ANXA1 FALSE TRUE
TRUE 1.08 7.85E-17 2966193 C6orf168 TRUE TRUE FALSE 1.37 1.01E-16
2525533 LOC648149 TRUE TRUE FALSE 1.24 1.02E-16 2525533 MAP2 TRUE
TRUE FALSE 1.24 1.02E-16 3154002 KCNQ3 TRUE TRUE FALSE 1.41
1.09E-16 3590164 SPINT1 TRUE TRUE FALSE 1.17 1.35E-16 3329343 MDK
TRUE TRUE TRUE 1.28 1.58E-16 2875193 P4HA2 TRUE TRUE FALSE 1.10
1.80E-16 3726691 ABCC3 TRUE TRUE FALSE 1.17 1.86E-16 2451870 ETNK2
TRUE TRUE TRUE 1.33 1.91E-16 4018327 TRPC5 TRUE TRUE TRUE 1.48
2.43E-16 3046197 ELMO1 TRUE TRUE TRUE -1.26 2.80E-16 2460817
SIPA1L2 TRUE TRUE TRUE 1.17 3.16E-16 3976341 TIMP1 TRUE TRUE TRUE
1.15 3.39E-16 2973232 C6orf174 TRUE TRUE FALSE 1.42 3.78E-16
2973232 KIAA0408 TRUE TRUE FALSE 1.42 3.78E-16 3417809 NAB2 TRUE
TRUE FALSE 1.25 5.50E-16 2751936 GALNT7 TRUE TRUE FALSE 1.17
5.95E-16 2648535 SGEF TRUE FALSE FALSE 1.16 1.33E-15 3759587
LOC100129115 TRUE TRUE FALSE 1.34 1.47E-15 3759587 PLCD3 TRUE TRUE
FALSE 1.34 1.47E-15 3994710 MAMLD1 FALSE TRUE FALSE 1.37 1.80E-15
3581221 AHNAK2 TRUE TRUE FALSE 1.31 2.29E-15 3259253 C10orf131
FALSE TRUE TRUE 1.01 4.17E-15 3259253 ENTPD1 FALSE TRUE TRUE 1.01
4.17E-15 2562435 EDNRB FALSE TRUE FALSE 1.37 5.28E-15 2562435 SFTPB
FALSE TRUE FALSE 1.37 5.28E-15 3489138 CYSLTR2 TRUE TRUE TRUE 1.30
5.69E-15 3002640 EGFR TRUE TRUE TRUE 1.11 8.20E-15 2578790 LRP1B
FALSE TRUE FALSE -0.95 1.06E-14 3768535 FAM20A FALSE TRUE FALSE
1.25 1.11E-14 3044129 GGCT TRUE TRUE FALSE 1.11 1.12E-14 2980449
IPCEF1 TRUE TRUE TRUE -1.14 1.29E-14 4018454 AMOT TRUE TRUE FALSE
1.34 1.47E-14 3763390 TMEM100 TRUE TRUE TRUE 1.40 2.44E-14 2740067
ANK2 FALSE TRUE TRUE -0.89 2.57E-14 3622934 MYEF2 TRUE TRUE TRUE
1.03 4.13E-14 3622934 SLC24A5 TRUE TRUE TRUE 1.03 4.13E-14 2414958
TACSTD2 FALSE TRUE FALSE 1.29 5.50E-14 3321150 ARNTL TRUE TRUE TRUE
1.18 7.68E-14 3464860 DUSP6 TRUE TRUE FALSE 1.10 1.17E-13 3464860
LOC100131490 TRUE TRUE FALSE 1.10 1.17E-13 3217242 GABBR2 TRUE TRUE
TRUE 1.21 1.22E-13 3110608 TM7SF4 TRUE TRUE TRUE 1.23 2.16E-13
3110395 RIMS2 TRUE TRUE FALSE 1.13 2.54E-13 3649714 C16orf45 TRUE
TRUE FALSE 1.10 7.74E-13 3867264 CA11 TRUE TRUE FALSE 1.05 8.23E-13
3832280 C19orf33 TRUE TRUE FALSE 1.20 8.77E-13 3832280 YIF1B TRUE
TRUE FALSE 1.20 8.77E-13 2452440 KLHDC8A TRUE TRUE FALSE 1.08
1.39E-12 2608469 ITPR1 TRUE TRUE TRUE -1.10 1.71E-12 3577612
SERPINA1 FALSE TRUE FALSE 0.96 2.24E-12 3577612 SERPINA2 FALSE TRUE
FALSE 0.96 2.24E-12 4015548 XKRX TRUE TRUE FALSE 1.12 2.68E-12
3451814 MAFG FALSE TRUE TRUE 1.04 2.91E-12 3451814 NELL2 FALSE TRUE
TRUE 1.04 2.91E-12 2734421 ARHGAP24 FALSE TRUE FALSE -1.05 3.17E-12
2816298 IQGAP2 TRUE TRUE FALSE -1.10 5.75E-12 2524301 NRP2 FALSE
TRUE FALSE 0.93 7.41E-12 3132616 ZMAT4 FALSE TRUE TRUE -0.89
1.03E-11 3365136 SERGEF FALSE TRUE TRUE 0.98 1.04E-11 3367673
MPPED2 FALSE TRUE FALSE -0.95 1.18E-11 2608309 LRRN1 FALSE FALSE
TRUE 0.84 1.66E-11 2820925 RHOBTB3 FALSE TRUE TRUE 0.85 2.73E-11
3369931 RAG2 FALSE TRUE TRUE -0.75 3.90E-11 2708922 IGF2BP2 FALSE
TRUE TRUE 0.90 5.15E-11 3868783 KLK7 TRUE TRUE TRUE 1.19 7.94E-11
3006572 AUTS2 TRUE TRUE FALSE 1.06 1.02E-10 3411810 PDZRN4 TRUE
TRUE FALSE 1.20 1.21E-10 2876897 SPOCK1 TRUE FALSE FALSE 1.05
1.39E-10 3166644 TMEM215 FALSE FALSE TRUE 0.98 1.49E-10 3933536
TFF3 FALSE TRUE FALSE -0.80 2.50E-10 3159330 DOCK8 FALSE TRUE TRUE
-0.90 2.53E-10 3279058 ACBD7 FALSE TRUE TRUE 1.03 2.83E-10 3593931
GLDN TRUE TRUE FALSE 1.13 3.46E-10 3404030 KLRG1 FALSE TRUE TRUE
-0.88 5.39E-10 2373842 PTPRC FALSE FALSE TRUE -0.90 9.75E-10
3010503 CD36 FALSE TRUE TRUE -0.81 3.46E-09 2583374 PLA2R1 FALSE
TRUE TRUE -0.72 6.14E-09 3856646 ZNF208 FALSE FALSE TRUE 0.77
6.91E-09 3692999 MT1G FALSE TRUE TRUE -0.82 1.01E-08 2587790 GPR155
FALSE TRUE FALSE -0.86 1.12E-08 2362351 PYHIN1 FALSE FALSE TRUE
-0.76 1.46E-08 2727587 KIT FALSE TRUE FALSE -0.75 1.50E-08 2427619
KCNA3 FALSE FALSE TRUE -0.78 1.50E-08 3142381 FABP4 FALSE TRUE
FALSE -0.72 1.82E-08 2584018 DPP4 FALSE TRUE TRUE 0.78 2.22E-08
2387126 RYR2 FALSE TRUE TRUE -0.64 2.26E-08 2823880 CAMK4 FALSE
FALSE TRUE -0.72 2.67E-08 3410384 C12orf35 FALSE FALSE TRUE -0.78
2.74E-08 2466554 TPO FALSE TRUE FALSE -0.77 5.30E-08 2806468 IL7R
FALSE FALSE TRUE -0.78 1.04E-07 2730746 SLC4A4 FALSE TRUE TRUE
-0.73 1.12E-07 3467949 SLC5A8 FALSE FALSE TRUE -0.74 1.23E-07
2518272 CERKL FALSE FALSE TRUE -0.74 1.58E-07 2518272 ITGA4 FALSE
FALSE TRUE -0.74 1.58E-07 3450861 ABCD2 FALSE FALSE TRUE -0.66
1.63E-07 3389450 CARD16 FALSE FALSE TRUE -0.78 1.66E-07 3389450
CASP1 FALSE FALSE TRUE -0.78 1.66E-07 2657831 IL1RAP FALSE TRUE
FALSE 0.78 1.85E-07 3059667 SEMA3D FALSE TRUE TRUE -0.71 2.04E-07
4013460 CYSLTR1 FALSE FALSE TRUE -0.71 2.12E-07 3126504 CSGALNACT1
FALSE TRUE TRUE -0.65 2.29E-07 3811339 BCL2 FALSE TRUE TRUE -0.76
2.29E-07 2724671 RHOH FALSE FALSE TRUE -0.69 2.37E-07 3160895 JAK2
FALSE FALSE TRUE -0.74 2.48E-07 2486811 PLEK FALSE FALSE TRUE -0.75
2.66E-07 3443804 KLRB1 FALSE FALSE TRUE -0.73 2.84E-07 3576704 TC2N
FALSE TRUE TRUE -0.74 3.29E-07 3742627 C17orf87 FALSE FALSE TRUE
-0.70 4.80E-07 3347658 ATM FALSE FALSE TRUE -0.65 4.89E-07 3347658
NPAT FALSE FALSE TRUE -0.65 4.89E-07 2815220 TMEM171 FALSE FALSE
TRUE -0.60 5.00E-07 3960174 LGALS2 FALSE FALSE TRUE -0.70 5.58E-07
2462329 ERO1LB FALSE TRUE TRUE -0.67 6.74E-07 2608725 BHLHE40 FALSE
TRUE TRUE 0.72 8.08E-07 3389353 CARD17 FALSE FALSE TRUE -0.72
1.09E-06 3389353 CASP1 FALSE FALSE TRUE -0.72 1.09E-06 3062082 PDK4
FALSE FALSE TRUE 0.67 1.22E-06 2593159 STK17B FALSE FALSE TRUE
-0.65 1.88E-06 2353669 CD2 FALSE FALSE TRUE -0.67 2.06E-06 2428796
PTPN22 FALSE FALSE TRUE -0.66 2.70E-06 2422035 GBP5 FALSE FALSE
TRUE -0.69 3.37E-06 2766289 TMEM156 FALSE FALSE TRUE -0.57 4.55E-06
3060450 C7orf62 FALSE FALSE TRUE -0.61 5.81E-06 2439554 AIM2 FALSE
FALSE TRUE -0.60 6.78E-06 3443891 CLEC2B FALSE FALSE TRUE -0.58
3.51E-05 2766192 TLR10 FALSE FALSE TRUE -0.51 3.87E-05 3536706
LGALS3 FALSE TRUE FALSE 0.52 4.67E-05 3009838 CCDC146 FALSE FALSE
TRUE -0.56 7.30E-05 3009838 POLR2J4 FALSE FALSE TRUE -0.56 7.30E-05
2412312 TTC39A FALSE FALSE TRUE 0.51 7.45E-05 2548699 CYP1B1 FALSE
TRUE FALSE 0.49 3.52E-04 3443868 CD69 FALSE FALSE TRUE -0.47
4.85E-04 3461981 TSPAN8 FALSE FALSE TRUE -0.44 7.33E-04 3648391
TNFRSF17 FALSE FALSE TRUE -0.44 7.66E-04 3018605 SLC26A4 FALSE TRUE
TRUE -0.46 9.81E-04 3107828 PLEKHF2 FALSE FALSE TRUE -0.42 1.19E-03
2372812 RGS13 FALSE FALSE TRUE -0.38 1.66E-03 3197955 GLDC FALSE
FALSE TRUE -0.37 5.51E-03 2796995 SORBS2 FALSE FALSE TRUE -0.32
1.01E-02 3135567 LYPLA1 FALSE FALSE TRUE -0.32 1.78E-02 2732508
CXCL13 FALSE FALSE TRUE -0.30 1.94E-02 3200982 MLLT3 FALSE FALSE
TRUE -0.30 2.03E-02 2735027 SPP1 FALSE FALSE TRUE 0.25 6.47E-02
2554018 EFEMP1 FALSE FALSE TRUE -0.20 1.55E-01 2945882 CMAH FALSE
FALSE TRUE -0.21 1.65E-01 2767378 ATP8A1 FALSE FALSE TRUE 0.20
1.79E-01 4016193 TMSB15A FALSE FALSE TRUE -0.16 2.27E-01 4016193
TMSB15B FALSE FALSE TRUE -0.16 2.27E-01 3019158 LRRN3 FALSE FALSE
TRUE 0.16 2.57E-01 2700244 CP FALSE FALSE TRUE 0.12 4.37E-01
2700244 HPS3 FALSE FALSE TRUE 0.12 4.37E-01 2855285 CCDC152 FALSE
FALSE TRUE -0.10 4.49E-01 2855285 SEPP1 FALSE FALSE TRUE -0.10
4.49E-01 2773947 CXCL9 FALSE FALSE TRUE -0.10 4.56E-01 3108226 PGCP
FALSE TRUE TRUE 0.04 7.65E-01 2773972 CXCL11 FALSE FALSE TRUE 0.02
8.93E-01
[0361] While preferred embodiments of the present invention have
been shown and described herein, it will be obvious to those
skilled in the art that such embodiments are provided by way of
example only. Numerous variations, changes, and substitutions will
now occur to those skilled in the art without departing from the
invention. It should be understood that various alternatives to the
embodiments of the invention described herein may be employed in
practicing the invention. It is intended that the following claims
define the scope of the invention and that methods and structures
within the scope of these claims and their equivalents be covered
thereby.
* * * * *
References