U.S. patent application number 11/282025 was filed with the patent office on 2006-06-29 for iterative probe design and detailed expression profiling with flexible in-situ synthesis arrays.
This patent application is currently assigned to Rosetta Inpharmatics LLC. Invention is credited to Julia Burchard, Stephen H. Friend, Peter S. Linsley, Roland Stoughton.
Application Number | 20060141501 11/282025 |
Document ID | / |
Family ID | 27386088 |
Filed Date | 2006-06-29 |
United States Patent
Application |
20060141501 |
Kind Code |
A1 |
Friend; Stephen H. ; et
al. |
June 29, 2006 |
Iterative probe design and detailed expression profiling with
flexible in-situ synthesis arrays
Abstract
Methods and compositions are provided that are useful for
detecting and reporting a plurality of different target
polynucleotide sequences in a sample, such as polynucleotides
corresponding to a plurality of different genes expressed by a cell
or cells. In particular, the invention provides methods for
screening a plurality of candidate polynucleotide probes to
evaluate both the sensitivity and the specificity with which each
candidate probe hybridizes to a target polynucleoide sequence.
Candidate polynucleotide probes can then be ranked according to
both their sensitivity and specificity, and probes that have
optimal sensitivity and specificity for a target polynucleotide
sequence can be selected. In one embodiment, polynucleotide probes
can be selected according to the methods described herein to
prepare "screening chips" wherein a large number of target
polynucleotide sequences are detected using a single microarray
have a few (e.g., 1-5) probes for each target polynucleotide
sequence. In a particularly preferred embodiment, the invention
provides a screening chip that can detect genetic transcripts from
the entire genome of an organism. In an alternative embodiment,
polynucleotide probes can be selected according to the methods
described herein to prepare "signature chips" to more accurately
detect certain selected "signature genes" using several
polynucleotide probes (e:g., 10-20) for each signature gene. The
invention additionally provides microarrays containing
polynucleotide probes for a large number of genes expressed by a
cell or organism. Further, methods for detecting a plurality of
polynucleotide molecules, including a large number of genes
expressed by a cell or organism, are also provided.
Inventors: |
Friend; Stephen H.;
(Seattle, WA) ; Stoughton; Roland; (San Diego,
CA) ; Linsley; Peter S.; (Seattle, WA) ;
Burchard; Julia; (Kirkland, WA) |
Correspondence
Address: |
JONES DAY
222 EAST 41ST ST
NEW YORK
NY
10017
US
|
Assignee: |
Rosetta Inpharmatics LLC
|
Family ID: |
27386088 |
Appl. No.: |
11/282025 |
Filed: |
November 16, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09561487 |
Apr 28, 2000 |
7013221 |
|
|
11282025 |
Nov 16, 2005 |
|
|
|
09364751 |
Jul 30, 1999 |
|
|
|
09561487 |
Apr 28, 2000 |
|
|
|
60144382 |
Jul 16, 1999 |
|
|
|
Current U.S.
Class: |
435/6.12 |
Current CPC
Class: |
G16B 25/00 20190201 |
Class at
Publication: |
435/006 |
International
Class: |
C12Q 1/68 20060101
C12Q001/68 |
Claims
1-212. (canceled)
213. A method for selecting one or more different polynucleotide
probes for detecting a target polynucleotide, said target
polynucleotide comprising a nucleotide sequence of a gene or gene
product expressed by a cell or organism, said one or more different
polynucleotide probes being selected from a plurality of different
candidate polynucleotide probes, each different candidate
polynucleotide probe comprising a different sequence that is
complementary and hybridizable to a nucleotide sequence of said
target polynucleotide, said method comprising: (a) selecting from
said plurality of different candidate polynucleotide probes a
plurality of polynucleotide probes that have a magnitude of binding
energy for hybridization to said target polynucleotide that is
above a selected threshold; (b) ranking the plurality of
polynucleotide probes selected in step (a) according to (b1) probe
length, and (b2) distance of the complementary sequence for each
probe in the target polynucleotide from one end of said target
polynucleotide; (c) de-overlapping said ranked polynucleotide
probes by a method which comprises (c1) selecting the top ranked
polynucleotide probe from said ranked polynucleotide probes, and
(c2) repeatedly selecting the next polynucleotide probe from said
ranked polynucleotide probes which comprises a nucleotide sequence
complementary to a sequence in said target polynucleotide that
overlaps the complementary sequence of the previously selected
polynucleotide probe in said target polynucleotide by no more than
a selected number of nucleotide bases; and (d) ranking said
de-overlapped polynucleotide probes according to the most negative
binding energy with which each of said de-overlapped polynucleotide
probes cross-hybridizes to one or more other polynucleotide
sequences expressed by said cell or organism, wherein said one or
more different polynucleotide probes for detecting said target
polynucleotide are selected from said ranked, de-overlapped
polynucleotide probes.
214. The method of claim 213 wherein the nucleotide sequences of
said target polynucleotide that are complementary to said candidate
polynucleotide probes are within a selected distance from one end
of said target polynucleotide.
215. The method of claim 213 further comprising, before said step
of selecting polynucleotide probes having a magnitude of binding
energy for hybridization to said target polynucleotide that is
above said selected threshold, steps of: (i) rejecting candidate
polynucleotide probes which comprise one or more sequences
corresponding to a repetitive element, a simple repeat or a polyX
repeat; and (ii) rejecting candidate polynucleotide probes having a
fraction of one or more selected nucleotide bases or a mathematical
combination of fractions of one or more particular nucleotide bases
which is not within a selected range of values.
216. The method of claim 213 wherein each said binding energy is
calculated according to a nearest neighbor model.
217. The method of claim 213 wherein said one or more other
polynucleotide sequences expressed by said cell or organism that
cross-hybridize to a candidate polynucleotide probe are identified
by a method which comprises identifying polynucleotide sequences
expressed by said cell or organism comprising nucleotide sequences
with a selected level of homology or identity to the complementary
sequence of said candidate polynucleotide probe.
218. The method of claim 213, wherein said step (b) is carried out
by ranking the plurality of polynucleotide probes selected in step
(a) according to a weighted combination of length and
end-distance.
219. A computer system for selecting one or more different
polynucleotide probes for detecting a target polynucleotide, said
computer system comprising: (a) a memory; and (b) a processor
element interconnected with the memory, wherein the memory encodes
one or more programs causing the processor to perform the method of
claim 213.
220. A computer program product for use in conjunction with a
computer having a memory and a processor, said computer program
product comprising a computer readable storage medium having a
computer program mechanism encoded thereon, wherein said computer
program mechanism may be loaded into the memory of a computer and
cause a processor of the computer to carry out the method of claim
213.
221. A method for selecting one or more different polynucleotide
probes for detecting a target polynucleotide, said target
polynucleotide comprising a nucleotide sequence of a gene or gene
product expressed by a cell or organism, said one or more different
polynucleotide probes being selected from a plurality of different
candidate polynucleotide probes, each different candidate
polynucleotide probe comprising a different sequence that is
complementary and hybridizable to a nucleotide sequence of said
target polynucleotide, said method comprising: (a) selecting from
said plurality of different candidate polynucleotide probes a
plurality of polynucleotide probes that have a magnitude of binding
energy for hybridization to said target polynucleotide that is
above a selected threshold; (b) ranking said plurality of
polynucleotide probes selected in step (a) according to a minimax
score, said minimax score for each said selected polynucleotide
probe (b1) being the largest magnitude of binding energy with which
said selected polynucleotide probes cross-hybridizes to one or more
other polynucleotide sequences expressed by said cell or organism;
or (b2) being determined using the product of the binding energy
with which said selected polynucleotide probes cross-hybridizes to
each of said one or more other polynucleotide sequences expressed
by said cell or organism and a weighting factor; and (c) selecting
one or more different polynucleotide probes from said plurality of
polynucleotide probes according to their ranks, thereby selecting
one or more different polynucleotide probes for detecting said
target polynucleotide.
222. The method of claim 221, further comprising prior to said step
(a) selecting said plurality of different candidate polynucleotide
probes such that said plurality of different candidate
polynucleotide probes are complementary to nucleotide sequences in
said target polynucleotide within a selected distance from one end
of said target polynucleotide.
223. The method of claim 222, wherein said end of said target
polynucleotide is the preferentially labeled end.
224. The method of claim 223, wherein said step (a) further
comprises the steps of (i) ranking said selected plurality of
polynucleotide probes according to distances of their respective
complementary sequences in the target polynucleotide from the
preferentially labeled end of said target polynucleotide; and (ii)
selecting one or more polynucleotide probes from said ranked
polynucleotide probes according to said rank.
225. The method of claim 221, further comprising before said step
(b) de-overlapping said plurality of polynucleotide probes by a
method comprising (i) ranking said plurality of polynucleotide
probes according to binding energy for hybridization to said target
polynucleotide; (ii) selecting the top ranked polynucleotide probe
from said ranked polynucleotide probes; (iii) repeatedly selecting
the next polynucleotide probe from said ranked polynucleotide
probes which is complementary to a sequence in said target
polynucleotide that overlaps the sequence complementary to the
sequence of the previously selected polynucleotide probe in said
target polynucleotide by no more than a selected number of
nucleotide bases; and (iv) rejecting polynucleotide probe or probes
not selected in step (iii) from said plurality of polynucleotide
probes.
226. The method of claim 221, further comprising prior to said step
(a) one or more of the following steps: (a1) rejecting candidate
polynucleotide probes which comprise one or more sequences
corresponding to a repetitive element, a simple repeat or a polyX
repeat; and (a2) rejecting candidate polynucleotide probes having a
fraction of one or more selected nucleotide bases or a mathematical
combination of fractions of one or more selected nucleotide bases
which is not within a selected range of values.
227. The method of claim 221, wherein said plurality of different
candidate polynucleotide probes comprises a set of successive
overlapping probes tiled along a sequence region of said target
polynucleotide at a tiling interval of between 1 and 10.
228. The method of claim 221, wherein said one or more other
polynucleotide sequences expressed by said cell or organism that
cross-hybridize to a candidate polynucleotide probe comprise
polynucleotide sequences expressed by said cell or organism
comprising nucleotide sequences with a selected level of homology
or identity to the complementary sequence of said candidate
polynucleotide probe.
229. The method of claim 221, wherein said selecting in step (c) is
carried out by selecting one or more different polynucleotide
probes that have the lowest minimax score or scores.
230. The method of claim 221, wherein each of said plurality of
different candidate polynucleotide probes consists of 40 to 70
nucleotides.
231. A computer system for selecting one or more different
polynucleotide probes for detecting a target polynucleotide, said
computer system comprising: (a) a memory; and (b) a processor
element interconnected with the memory, wherein the memory encodes
one or more programs causing the processor to perform the method of
claim 22 1.
232. A computer program product for use in conjunction with a
computer having a memory and a processor, said computer program
product comprising a computer readable storage medium having a
computer program mechanism encoded thereon, wherein said computer
program mechanism may be loaded into the memory of a computer and
cause a processor of the computer to carry out the method of claim
221.
Description
[0001] This is a divisional of copending U.S. patent application
Ser. No. 09/561,487, filed on Apr. 28, 2000, which is a
continuation-in-part application of U.S. patent application Ser.
No. 09/364,751, filed on Jul. 30, 1999, now abandoned, which claims
benefit under 35 U.S.C. .sctn. 119(e) of U.S. Provisional Patent
Application Ser. No. 60/144,382, filed on Jul. 16, 1999, each of
which is incorporated herein by reference in its entirety.
1. FIELD OF THE INVENTION
[0002] The field of this invention relates to materials and methods
to detect and report polynucleotide sequences, including genomic
sequences, genomic transcript sequences (e.g., mRNAs from cells
and/or cDNA sequences derived therefrom), copy numbers and SNPs. In
particular, the invention relates to methods for detecting
polynucleotide sequences using sets of polynucleotide probes that
have been selected for optimum sensitivity and specificity. The
invention also relates to methods for selecting sets of
polynucleotide probes for optimum sensitivity and specificity which
may be used, e.g., to detect and report gene expression changes in
a cell or cells. The invention further relates to sets of
polynucleotide probes, including microarrays comprising such sets
of polynucleotide probes, which are selected for optimum
sensitivity and specificity and are therefore useful, e.g., to
detect and report gene expression changes in a cell or cells.
2. BACKGROUND
[0003] Within the past decade, several technologies have made it
possible to monitor the expression level of a large number of
genetic transcripts at any one time (see, e.g., Schena et al.,
1995, Science 270:467-470; Lockhart et al., 1996, Nature
Biotechnology 14:1675-1680; Blanchard et al., 1996, Nature
Biotechnology 14:1649; Ashby et al., U.S. Pat. No. 5,569,588,
issued Oct. 29, 1996). For example, techniques are known for
preparing microarrys of cDNA transcripts (see, e.g., DeRisi et al.,
1996, Nature Genetics 14:457-460; Shalon et al., 1996, Genome Res.
6:689-645; and Schena et al., 1995, Proc. Natl. Acad. Sci. U.S.A.
93:10539-11286). Alternatively, high-density arrays containing
thousand of oligonucleotides complementary to defined sequences, at
defined locations on a surface using photolithographic techniques
for synthesis in situ are described, e.g., Fodor et al., 1991,
Science 251:767-773; Pease et al., 1994, Proc. Natl. Acad. Sci.
U.S.A. 91:5022-5026; Lockhart et al., 1996, Nature Biotechnology
14:1675; U.S. Pat. Nos. 5,578,832; 5,556,752;l and 5,510,270).
Methods for generating arrays using inkjet technology for
oligonucleotide synthesis are also known in the art (see, e.g.,
Blanchard, International Patent Publication WO 98/41531, published
Sep. 24, 1998; Blanchard et al., 1996, Biosensors and
Bioelectronics 11:687-690; Blanchard, 1998, in Synthetic DNA Arrays
in Genetic Engineering, Vol. 20, J. K. Setlow, Ed., Plenum Press,
New York at pages 111-123).
[0004] Applications of this technology include, for example,
identification of genes which are up regulated or down regulated in
various physiological states, particularly diseased states.
Additional exemplary uses for transcript arrays include the
analyses of members of signaling pathways, and the identification
of targets for various drugs. See, e.g., Friend and Hartwell,
International Publication No. WO 98/38329 (published Sep. 3, 1998);
Stoughton, U.S. Pat. No. 6,132,969; Stoughton and Friend, U.S. Pat.
No. 5,965,352; Friend and Stoughton, U.S. Provisional Application
Ser. Nos. 60/084,742 (filed May 8, 1998), 60/090,004 (filed Jun.
19, 1998), and 60/090,046 (filed Jun. 19, 1998).
[0005] However, several factors limit the number of genetic
transcripts that can be detected on a single microarray "chip." In
particular, the "reporting density" (i.e., the number of genes
detected per unit of surface area) for a microarray is limited,
e.g., by the density with which polynucleotide probes may be laid
down as well as by the number of polynucleotide probes required per
gene. A plurality of probe pairs, which are both matched to and
intentionally mismatched to a target sequence, are required in
order to empirically distinguish signal arising from a target
polynucleotide sequence of interest (e.g., a particular mRNA
sequence of interest) from signal arising from cross-hybridization
with other polynucleotide sequences. Currently, in situ synthesized
microarray chips require more than 20 oligonucleotide probe pairs
per gene or gene region reported (Lockhart et al.,supra). On the
other hand, the number of polynucleotide probes that may be laid
down on a microarray chip is limited by the technology used to
produce the microarray. Photolithographic techniques discussed
above for producing oligonucleotide microarrays having a high
spatial density of probes are expensive to synthesize and therefore
require a large capital investment. Oligonucleotide microarrays
produced using the above discussed inkjet technology methods are,
by contrast, much cheaper and faster to produce both per chip
design and per chip. Thus, such microarrays are generally preferred
for detecting genetic transcripts in cells. However, microarray
chips produced by such inkjet technology have a limited probe
density that is only a fraction of the probe density of chips
produced by photolithography methods. Thus, at present the number
of genetic transcripts that may be detected on a single microarray
chip is limited to about 10,000 gene transcripts using expensive,
photolithographic arrays, and only about 750 to 2,500 gene
transcripts using less expensive, inkjet arrays.
[0006] There exists therefore a need for materials and methods
which may be used to efficiently detect large numbers of different
genetic transcripts and thereby detect changes in a large number of
genetic transcripts in a cell or cells. In particular, there is a
need for materials and methods which may be used to detect changes
in genetic transcription across the entire genome of a cell,
including cells of complex organisms such as mammalian cells and,
in particular, human cells.
[0007] There also exists, however, a need for materials and methods
which may be used to accurately detect changes in genetic
transcripts in cells, e.g., in response to some environmental
change or perturbation. In particular, there is a need to
accurately detect changes in the expression levels of those
particular genetic transcripts that exhibit the largest changes,
e.g., in response to an environmental change or perturbation, and
which are therefore most relevant in understanding the effect of
the environmental change or perturbation on the cell or cells.
[0008] Discussion or citation of a reference herein shall not be
construed as an admission that such reference is prior art to the
present invention.
3. SUMMARY OF THE INVENTION
[0009] The present invention provides methods and compositions that
efficiently detect and accurately report gene expression changes in
an organism. In particular, the methods and compositions of the
invention may be used to detect and report gene expression changes
in a cell or organism that occur, e.g., in response to some change
or "perturbation" to the cell or organism and/or to its
environment, such as exposure of the cell or organism to one or
more drugs.
[0010] The compositions and methods of the invention use "screening
chips," which may be used, e.g., to detect changes in gene
expression among a large number of genes or gene transcripts. For
example, in particularly preferred embodiments, the screening chips
may be used to detect changes in gene expression in the entire
genome of an organism. Such screening chips are therefore provided
as part of the present invention, as well as methods for making and
using such screening chips, e.g., to screen the entire genome of an
organism for changes in response to one or more perturbations.
[0011] The compositions and methods of the invention also provide
"signature chips" which may be used to accurately detect changes in
gene expression in a smaller number of genes. For example, the
signature chips of the invention may be used to accurately detect
changes in the expression of certain "signature genes." In
preferred embodiments, the signature genes are those genes whose
expression changes the most in response to a particular
perturbation or in response to a particular type or set of
perturbations (e.g., responses to several doses of a drug or
responses to several different, but related drugs). For example,
signature genes may be identified using the screening chips of the
invention to identify those genes whose expression changes the most
in response to a particular perturbation or perturbations. In one
preferred embodiment, the signature chips of the invention comprise
at least a first probe and a second probe for each signature gene
to be detected, wherein the first probe for a particular signature
gene is a matched probe having a polynucleotide sequence that is
complementary to the particular signature gene or to a portion
thereof, and wherein the second probe for a particular signature
gene is a mismatch probe having a polynucleotide sequence that is a
variant of a sequence which is complementary to the particular
signature gene. In another preferred embodiment the signature chips
of the invention comprise a plurality of matched probes for each
signature gene to be detected, wherein each matched probe for a
particular signature gene has a polynucleotide sequence that is
complementary to the particular signature gene or to a portion
thereof.
[0012] The invention also provides methods and compositions for
ranking and/or selecting probes according to other parameters
including, but not limited to: (a) probe size or length; (b)
binding energies, including both the perfect match duplex (i.e., of
a probe and its target, complementary nucleotide sequence) and
cross-hybridization binding energies; (c) base composition,
including, for example, the relative amount or percentage of one or
more particular nucleotide bases (e.g., adenine, guanine, thymine
or cytosine) in a probe sequence, as well as the relative amount or
percentage of any combination of such nucleotide bases; (d) the
position of a probe's complementary sequence in the sequence of its
"target" polynucleotide or gene sequence; and (e) probe sequence
complexity, including the presence or lack of common repetitive
elements such as polynucleotide repeats (i.e., simple, contiguous
repeats of one or more nucleotide bases) as well as more
complicated repetitive elements that are well known in the art.
Still other exemplary parameters which can be used in the methods
and compositions of the invention for ranking and/or selecting
oligonucleotide probes include: (f) self dimer binding energy
(i.e., the tendency for a particular probe to hybridize to its own
sequence); (g) the structure content of the complementary, target
polynucleotide sequence for a particular probe (e.g., the presence
or absence of certain structural features or motifs); and (h) the
information content of a probe's nucleotide sequence.
[0013] The invention is based, at least in part, on the discovery
that the number of probe sequences required to reliably and
accurately report a particular polynucleotide sequence, such as the
sequence of a particular gene, may be reduced to as few as one
probe by carefully selecting probes according to the methods and/or
having the particular lengths disclosed herein. Accordingly, the
invention also provides methods by which probes (i.e., probe
sequences) may be ranked and/or selected according to their
reporting properties, including, for example, their specificity and
sensitivity for a particular sequence (e.g., for the sequence of a
particular gene or gene transcript).
[0014] The invention thus provides methods for selecting one or
more different polynucleotide probes from a plurality of
polynucleotide probes according to the sensitivity and specificity
with which each different polynucleotide probe hybridizes to a
target polynucleotide. In one embodiment, the methods comprise: (a)
identifying polynucleotide probes in the plurality of different
polynucleotide probes that hybridize to the target polynucleotide
with a sensitivity above a threshold sensitivity level; (b) ranking
the identified polynucleotide probes according to the specificity
with which each identified polynucleotide probe hybridizes to the
target polynucleotide; and (c) selecting one or more different
polynucleotide probes from the ranked polynucleotide probes. In
another embodiment, the methods comprise: (a) identifying
polynucleotide probes in the plurality of different polynucleotide
probes that hybridize to the target polynucleotide with a
specificity above a threshold specificity level; (b) ranking the
identified polynucleotide probes according to the sensitivity with
which each identified polynucleotide probe hybridizes to the target
polynucleotide; and (c) selecting one or more different
polynucleotide probes from the ranked polynucleotide probes. In
still another embodiment, the methods comprise: (a) ranking the
plurality of different polynucleotide probes according to the
sensitivity with which each polynucleotide probe hybridizes to the
target polynucleotide so that a sensitivity rank is obtained for
each different polynucleotide probe; (b) ranking the plurality of
different polynucleotide probes according to the specificity with
which each polynucleotide probe hybridizes to the target
polynucleotide so that a specificity rank is obtained for each
different polynucleotide probe; (c) obtaining a combined rank for
each different polynucleotide probe, wherein the combined rank is
determined by determining the sum of the sensitivity rank and the
specificity rank for each different polynucleotide probe; and (d)
selecting one or more different polynucleotide probes from the
plurality of different polynucleotide probes according to the
combined rank of the different polynucleotide probes. In one aspect
of this particular embodiment, the sum of the sensitivity rank and
the specificity rank for each different polynucleotide probe can
be, e.g., a weighted sum of the sensitivity rank and the
specificity rank for each different polynucleotide probe.
[0015] The invention provides numerous different aspects of these
different embodiments for example, the invention provides aspects
of the above embodiments wherein the sensitivity with which a
particular polynucleotide probe hybridizes to the target is
provided by determining the binding energy with which the target
polynucleotide hybridizes to the particular polynucleotide probe,
e.g., according to the nearest neighbor model. The invention also
provides aspects of the above embodiments wherein the sensitivity
with which a particular polynucleotide probe hybridizes to the
target polynucleotide is provided by a method comprising
determining the level of hybridization of the target polynucleotide
sequence to the particular polynucleotide probe; e.g., by
calculating the level of hybridization of the target polynucleotide
to the polynucleotide probe from the binding energy with which the
target polynucleotide hybridizes to the particular polynucleotide
probe.
[0016] In another aspect of the methods of the invention, the
specificity with which a particular polynucleotide probe hybridizes
to the target polynucleotide is provided, e.g., by: (a) determining
the level of hybridization of the target polynucleotide to the
particular polynucleotide probe; and (b) determining the level of
cross-hybridization of non-target polynucleotides to the particular
probe.
[0017] In still other embodiments, the methods of the invention
comprise: (a) hybridizing a reference polynucleotide sample
comprising molecules of the target polynucleotide to the plurality
of different polynucleotide probes under conditions such that the
hybridization intensity of each different polynucleotide probe to
the reference sample correlates with the sensitivity and
specificity with which the each different polynucleotide probe
hybridizes to the target polynucleotide; and (b) selecting
polynucleotide probes in the plurality of different polynucleotide
probes that have the highest hybridization intensity. For example,
the invention provides particular aspects of this embodiment
wherein the hybridization is within 5.degree. C. or within
2.degree. C. of the mean melting temperature of the plurality of
different polynucleotide probes from the target polynucleotide.
[0018] The invention also provides a preferred embodiment wherein
the specificity of a particular polynucleotide probe is provided by
a method which comprises selecting, from a plurality of binding
energies, a binding energy that indicates the specificity of the
particular polynucleotide probe. Specifically, in such a preferred
embodiment, the provided plurality of binding energies are binding
energies for hybridization of the particular polynucleotide probe
to each of a plurality of different polynucleotides, wherein each
polynucleotide in the plurality of different polynucleotides is
different from the target polynucleotide. The selected binding
energy is the largest binding energy in the plurality of binding
energies.
[0019] For example, in one aspect of this preferred embodiment, the
binding energies provided for hybridization of the particular
polynucleotide probe to each of the plurality of polynucleotides is
provided according to a nearest neighbor model. In one aspect the
plurality of polynucleotides comprise polynucleotides expressed by
a cell or organism of interest. In one aspect, the plurality of
polynucleotides consists of polynucleotides having sequences with a
selected level of identity or homology to a complementary sequence
of the particular polynucleotide probe. For example, in one aspect,
the sequences having the selected level of identity or homology to
the complementary sequence of the probe are identified by means of
a BLAST or PowerBLAST algorithm. In various aspects, the plurality
of polynucleotides consists of polynucleotides having sequences
that are at least 50%, at least 60%, at least 70%, at least 80%, at
least 90%, at least 95% or at least 99% identical to the
complementary sequence of the particular polynucleotide probe.
[0020] In still other embodiments, which are both more general and
more preferred embodiments, the polynucleotide or oligonucleotide
probes are ranked and/or selected according to a combination of two
or more of the properties (a)-(h) listed above and, optionally, the
sensitivity and/or specificity with which each probe hybridizes to
a target polynucleotide. For example, in one embodiment the
invention provides methods for selecting one or more different
polynucleotide probes from a plurality of polynucleotide probes be
a method comprising: (a) identifying those polynucleotide probes in
the plurality of polynucleotide probes that have particular values
(or a particular range of values) of one, two, three or more
properties or parameters (e.g., selected among the properties and
parameters listed hereinabove); and (b) selecting the
polynucleotide probes identified in step (a).
[0021] In another general embodiment, the methods of the invention
comprise: (a) ranking the polynucleotide probes in a plurality of
different polynucleotide probes according to each of two or more
selected properties or parameters (e.g., selected from the
properties and parameters recited hereinabove) so that a rank is
obtained for each of the two or more selected parameters; and (b)
obtaining a combined rank for each different polynucleotide probe,
wherein the combined rank is determined from the sum of the ranks
obtained for each of the two or more selected properties or
parameters. One or more different polynucleotide probes can then be
selected from the plurality of different polynucleotide probes
according to the combined rank of the different polynucleotide
probes.
[0022] In yet another general embodiment, the methods of the
invention comprise: (a) identifying those polynucleotide probes in
the plurality of polynucleotide probes that have particular values
(or a particular range of values) of one, two, three or more
properties or parameters (e.g., selected among the properties and
parameters listed hereinabove); (b) ranking the identified
polynucleotide probes according to each of two or more selected
properties or parameters (e.g., selected among the properties and
parameters listed hereinabove) so that a rank is obtained for each
of the two or more selected parameters; and (c) obtaining a
combined rank for each identified polynucleotide probe, wherein the
combined rank is determined from the sum of the ranks obtained for
each of the two or more selected properties or parameters. One or
more different polynucleotide probes can then be selected from the
identified polynucleotide probes according to the combined rank of
the identified polynucleotide probes.
[0023] In such a general embodiment, the properties or parameters
used to rank the identified probes in step (b) can be either the
same as or, more preferably, different from the properties or
parameters used to identify those polynucleotide probes in step
(a). Also, in certain aspects of embodiments such as the general
embodiments described above, the sum of the ranks obtained for each
of the two or more selected properties or parameters can be, e.g.,
a weighted sum of the ranks obtained for each of the two or more
selected properties or parameters.
[0024] The invention provides certain preferred aspects of the
above methods wherein the steps of the methods are iteratively
repeated, e.g., to select no more than 20, 10, 5 or 1 different
polynucleotide probe or probes. The invention also provides
preferred aspects of these methods wherein the polynucleotide
probes comprise polynucleotide sequences that are, e.g., between
15-500, 20-100 or 40-60 bases in length.
[0025] The invention also provides, in still other embodiments,
screening chips and signature chips that comprise arrays of
polynucleotide probes selected according to the methods of the
invention. Specifically, the screening chips of the invention
comprise an array of a plurality of different polynucleotide probes
for a plurality of different target polynucleotides, wherein each
different polynucleotide probe in the plurality of different
polynucleotide probes is selected by any one of the above described
methods. In preferred embodiments, the screening chips comprise,
e.g., at least 4000, 10000, 15000, 20000, 80000, or 100000
different polynucleotide sequences. In other preferred embodiments,
the screening chips of the invention comprise no more than 10, 2 or
1 different polynucleotide probes that hybridize to a particular
target polynucleotide.
[0026] In yet other embodiments, the screening chips comprise an
array of a plurality of different polynucleotide probes for a
plurality of different target polynucleotides, wherein each
different polynucleotide probe is selected according to any one of
the methods of the invention and wherein the plurality of different
target polynucleotides comprise polynucleotide sequences of, e.g.,
at least 50%, 75%, 80%, 85%, 90%, 95%, 99% or 100% (i.e., all) of
the genes in the genome of a cell or organism; including particular
embodiments wherein the cell or organism is a human cell or
organism.
[0027] The signature chips of the invention comprise an array of a
plurality of different polynucleotide probes for one or more target
polynucleotides, wherein each different polynucleotide probe is
selected by one of the methods of the invention. In preferred
embodiments, the target polynucleotides comprise one or more
signature genes which comprise one or more genetic transcripts of a
cell or organism whose abundances change in response to one or more
changes or perturbations to the cell or organism.
[0028] In one preferred embodiment, the signature chips of the
invention comprise, for each target polynucleotide, at least one
pair of polynucleotide probes wherein each pair comprises: (a) a
match probe that is complementary to a particular target
polynucleotide; and (b) an intentional mismatch probe that differs
from the match probe in at least one nucleotide. In another
preferred embodiment, the signature chips of the invention
comprise, for each target polynucleotide, at least one set of
polynucleotide probes, with each set comprising: (a) a match probe
that is complementary to a particular target polynucleotide; and
(b) a plurality of (for example between 4 and 20) different
intentional mismatch probes which differ from the match probe in at
least one nucleotide.
[0029] The invention also provides, in still other embodiments,
methods for preparing signature chips comprising an array of
polynucleotide probes for one or more signature genes, wherein the
methods comprise: (a) identifying one or more target
polynucleotides corresponding to gene transcripts of a cell or
organism that change expression or abundances in response to one or
more particular changes or perturbations to the cell or organism,
said one or more target polynucleotides being said one or more
signature genes; (b) selecting a plurality of different
polynucleotide probes for each of said one or more signature genes
from a plurality of candidate polynucleotide probes according to
the sensitivity and specificity with which each candidate
polynucleotide probe hybridizes to one of said signature genes; and
(c) preparing a microarray comprising an array of the selected
polynucleotide probes for each of said one or more signature genes,
wherein said microarray is a signature chip. In one preferred
aspect of this embodiment, the one or more particular target
polynucleotides are identified using a screening chip, wherein the
screening chip comprises an array of different polynucleotide
probes for a plurality of different target polynucleotides, and
wherein each different polynucleotide probe of said screening chip
is selected according to the sensitivity and specificity with which
each different polynucleotide probe hybridizes to one of said
plurality of target polynucleotides.
[0030] In yet other embodiments, the invention further provides
arrays of polynucleotide probes. The arrays comprise a support with
at least one surface and at least 100 different polynucleotide
probes, each different polynucleotide probes comprising a different
polynucleotide sequence and being attached to the surface of the
support in a different location on the surface. The nucleotide
sequence of the different polynucleotide probes is in the range of
40 to 80 nucleotides in length, and in preferred embodiments is in
the range of 50 to 70, or 50 to 60 nucleotides in length. In
preferred aspects of this embodiment, the arrays comprise
polynucleotide probes of at least 4000, 10000, 15000, 20000, 50000,
80000, or 100000 different nucleotide sequences. Preferably, each
polynucleotide probe on the array is specific for a particular
target polynucleotide sequence. More preferably, the nucleotide
sequence of each different polynucleotide probe of the array is
specific for a different target polynucleotide sequence.
Preferably, the target polynucleotide sequences comprise expressed
polynucleotide sequences of a cell or organism, such as a mammalian
cell or organism (e.g., a human cell or organism), and the
nucleotide sequences of the different probes of the array are
specific for at least 50%, 75%, 80%, 85%, 90%, 95%, 99% or 100%
(i.e., all) of the genes in the genome of the cell or organism.
[0031] In still other preferred embodiments, the arrays comprise at
least 100, at least 1000, or at least 2500 different probes per 1
cm.sup.2. In other preferred embodiments the array is a
positionally addressable array. In yet other preferred embodiments,
the different polynucleotide probes comprise sets of polynucleotide
probes, each set of polynucleotide probes comprising: (a) a match
probe having a nucleotide sequence that is complementary to a
particular target polynucleotide sequence, and (b) at least one
intentional mismatch probe having a nucleotide sequence which
differs from the nucleotide sequence of the match probe in at least
one nucleotide, and more preferably in one to three
nucleotides.
[0032] The invention still further provides, in other embodiments,
systems (e.g., computer systems) for executing the methods of the
invention. In particular, a computer system of the invention
comprises a memory and a processor interconnected with the memory,
wherein the memory encodes one or more programs causing the
processor to perform one or more of the above-related methods. The
invention also provides computer program products for using in
conjunction with a computer having a memory and a processor. The
computer program products of the invention comprise a computer
readable storage medium having a computer program mechanism encoded
thereon, wherein the computer program mechanism may be loaded into
the memory of a computer and causes a processor of the computer to
execute the steps of one or more of the above-recited methods.
[0033] In addition, the present invention also provides methods for
detecting whether a plurality of polynucleotide molecules is
present in a sample. Such methods comprise steps of: (a) contacting
a sample comprising polynucleotide molecules to an array of the
invention under conditions that permit the polynucleotide molecules
in the sample to hybridize to the array; and (b) detecting any
hybridization of polynucleotide molecules in the sample to
polynucleotide probes of the array. Hybridization of a particular
polynucleotide molecules to a polynucleotide probe of the array
then indicates the presence of that particular polynucleotide
molecule in the sample. In particularly preferred aspects of this
embodiment, the methods are used to detect polynucleotides
expressed by a cell or organism (e.g., expressed by at least 50%,
75%, 85%, 90%, 95% or 99% of the expressed genes in the genome of
the cell or organism). In particular, in such preferred aspects the
sample comprises polynucleotide molecules, such as mRNA molecules,
expressed by the cell or organism, or polynucleotide molecules such
as cDNA molecules or cRNA molecules that are derived therefrom.
[0034] It yet another embodiment, the invention also provides
methods for detecting whether a plurality of polynucleotide
molecules is present in a sample. The methods comprise: (a)
contacting a sample comprise polynucleotide molecules to an array
under conditions that permit polynucleotide molecules in said
sample to hybridize to polynucleotide probes of said array; and (b)
detecting any hybridization of polynucleotide molecules in the
sample to polynucleotide probes of the array. Hybridization of a
particular polynucleotide molecule to a polynucleotide probe of the
array indicates the presence of the particular polynucleotide
molecule in the sample. In preferred aspects of this embodiment,
the array comprises a support with at least one surface and having
at least 100 different polynucleotide probes. Each different
polynucleotide probe: (i) comprises a different nucleotide
sequence, (ii) is attached to the surface of the support at a
different location on the surface, and (iii) has a nucleotide
sequence 40 to 80 nucleotides in length.
4. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] FIG. 1 provides a flow chart illustrating an exemplary
embodiment of the general methods of the present invention.
[0036] FIG. 2 depicts the predicted "melting curve," i.e., the
fraction of target polynucleotide molecules bound to an
oligonucleotide probe as a function of the hybridization
temperature, for perfect match polynucleotide molecules (PM), one
base mismatch polynucleotide molecules (1MM), and two base mismatch
polynucleotide molecules (2MM); also depicted as a function of
temperature are the hybridization ratios of perfect match
polynucleotide molecules to 1 base mismatch polynucleotide
molecules (Ratio PM/1MM) and of perfect match polynucleotide
molecules to 2 base mismatch polynucleotide molecules (Ratio
PM/2MM).
[0037] FIGS. 3A-D show experimental data demonstrating that
hybridization specificity of a collection of probes is optimized at
or slightly above the mean melting temperature of the probes;
specifically FIG. 3A shows the mean hybridization intensity ratio
of perfect-match to single mismatch (PM/SM,--) and perfect-match to
double mismatch (PM/DM,--) vs. the hybridization temperature
observed for a collection of 22-mer oligonucleotide probes; FIG. 3B
is a histogram showing the distribution of perfect match melting
temperatures (T.sub.m) predicted for the 22-mer probes; FIG. 3C
shows the mean hybridization intensity ratio of perfect-match to
single mismatch (PM/SM,--) and perfect-match to double mismatch
(PM/DM,--) vs. the hybridization temperature observed for a
collection of 35-mer oligonucleotide probes; FIG. 3D is a histogram
showing the distribution of perfect match melting temperatures
(T.sub.m) predicted for the 35-mer probes.
[0038] FIG. 4 plots the schematic behavior of the intensity and
specificity of hybridization for polynucleotide probes as a
function of their binding energies .DELTA.G.
[0039] FIGS. 5A-C show the amount of target and non-targeted
hybridization observed for individual probes targeted for the S.
cerevisiae gene YER019W, individual probes are identified according
to their "tiling position"; FIG. 5A plots the mean normalized
hybridization intensity for a polynucleotide sample which contains
only YER019W polynucleotides ("targeted hybridization"); FIG. 5B
plots the mean normalized hybridization intensity for a
polynuicleotide sample derived from an S. cerevisiae strain deleted
for the gene YER019W ("non-targeted hybridization"); FIG. 5C plots
the ratio of targeted to non-targeted hybridization intensities
shown in FIGS. 5A-B; those probes which are predicted to have the
highest specificity are marked with an (X) symbol.
[0040] FIGS. 6A-C show the amount of target and non-targeted
hybridization observed for individual probes targeted for the S.
cerevisiae gene HXT3, individual probes are identified according to
their "tiling position"; FIG. 6A plots the mean normalized
hybridization intensity for a polynucleotide sample which contains
only HXT3 polynucleotides ("targeted hybridization"); FIG. 6B plots
the mean normalized hybridization intensity for a polynucleotide
sample derived from an S. cerevisiae strain deleted for the gene
HXT3 ("non-targeted hybridization"); FIG. 6C plots the ratio of
targeted to non-targeted hybridization intensities shown in FIGS.
6A-B; those probes which are predicted to have the highest
specificity are marked with an (X) symbol.
[0041] FIGS. 7A-B show plots of binding energy and specificity for
a plurality of oligonucleotide probes to the S. cerevisiae genes
YER019W and YAR010C using binding energy and specificity values
calculated according to the methods described hereinbelow; FIG. 7A
is a plot of the energy score (i.e., the binding energy) vs. the
cross-hybridization score (i.e., the specificity) of probes to the
gene YER019W; FIG. 7B is a plot of the energy score (i.e., the
binding energy) vs. the cross-hybridization score (i.e., the
specificity) of probes to the gene YAR010C.
[0042] FIGS. 8A-B show plots of hybridization intensity vs.
specificity for a plurality of oligonucleotide probes to the S.
cerevisiae genes YER019W and HXT3, using the experimental data
displayed in FIGS. 5 and 6; FIG. 8A plots the observed
hybridization intensities vs. specificity for oligonucleotide
probes to the gene YER019W; FIG. 8B plots the observed
hybridization intensities vs. specificity for oligonucleotide
probes to the gene HXT3.
[0043] FIG. 9 is a representation of changes in abundances of 4,000
gene transcripts of S. cerevisiae as a result of 350 different
changes or perturbations to cells.
[0044] FIG. 10 is a representation of a computer system which may
be used to practice the analytical methods of the present
invention.
[0045] FIG. 11 is a distribution plot comparing expression ratios
measured using a screening chip of the invention (horizontal axis)
and a standard microarray (vertical axis).
[0046] FIG. 12 shows a histogram of the distribution of fractional
errors for the absolute hybridization intensities (dashed line) and
expression ratios (solid line) from hybridization data measured
using a screening chip of the invention.
[0047] FIG. 13 is a plot showing the specificity of oligonucleotide
probes for both sequence length (vertical axis) and hybridization
stringency (horizontal axis).
[0048] FIGS. 14A-C are scatter plots comparing the changes in
"signature" genes in samples of RNA from unactivated to activated
human lymphocytes; the horizontal axis indicates changes measured
using a signature chip of the invention with an average of 17
oligonucleotide probes per gene; the vertical axis indicates
changes measured using a screening chip with only one
oligonucleotide probe per gene; FIG. 14A is a scatter plot
comparing data for 164 genes for which significant changes were
detected with both a screening chip of the invention and a
traditional "spotter chip;" FIG. 14B is a scatter plot comparing
data for 237 genes for which significant changes were observed with
a screening chip but not on a spotter chip; FIG. 14C is a scatter
plot comparing data for 149 genes for which significant changes
were not observed with a screening chip but were observed with a
spotter chip.
[0049] FIGS. 15A-D are exemplary signature plots from four
signature genes of the 149 depicted in FIG. 14C for which
significant changes were observed in experiments using traditional
spotter chips but not in experiments using screening chips; the
genes were categorized into four separate classes; FIG. 15A is an
exemplary signature plot of a gene (L11066) in Class 1; FIG. 15B is
an exemplary signature plot of a gene (M76541) in Class 2; FIG. 15C
is an exemplary signature plot of a gene (U33017) in Class 3; FIG.
15D is an exemplary signature plot of a gene (X1 7620) in Class
4.
[0050] FIG. 16 shows the ratio of ETR103 expression in activated
and unactivated human lymphoblast cells reported by exemplary
candidate oligonucleotide probes (vertical axis), plotted against
the fraction of guanine (G) and cytosine (C) nucleotide bases in
each probe.
[0051] FIG. 17 shows a plot of the ratio of AML1b expression in
Jurkat to K562 cells reported by exemplary candidate
oligonucleotide probes (vertical axis) verses the position of the
each probe's complementary sequence in the AML1b gene.
[0052] FIGS. 18A-B illustrate the effect of simple and complex
repetitive sequence elements on oligonucleotide probe specificity;
FIG. 18A plots the reported differential hybridization of exemplary
candidate probes to the ERT103 gene (vertical axis) plotted against
the hybridization intensity of each probe (horizontal axis) with
probes containing one or more of the repetitive elements
(CAG).sub.n, (CGG).sub.n and (AGGGGG).sub.n indicated by open
circles; FIG. 18B plots the reported differential hybridization of
exemplary candidate probes to the AIM1 gene (vertical axis) plotted
against the hybridization intensity of each probe (horizontal axis)
with probes for which greater than 60% of the probe sequence is
contained within an ALU repeat identified in the AML1b gene being
indicated by open circles.
[0053] FIG. 19 shows a flow chart illustrating a preferred,
exemplary embodiment of the ranking methods of the invention.
5. DETAILED DESCRIPTION
[0054] The present invention provides methods and compositions for
detecting and reporting changes in gene expression in a cell or
cells. In particular, the invention provides methods and
compositions that may be used to efficiently and accurately detect
a plurality of target polynucleotides in a sample, e.g., by
hybridization to a microarray. The invention therefore relates to
hybridization of samples comprising a plurality of different target
polynucleotides to a plurality of different probes for those target
polynucleotides.
[0055] Exemplary target polynucleotides which may be analyzed by
the methods and compositions of the present invention include, but
are not limited to DNA molecules such as genomic DNA molecules,
cDNA molecules, and fragments thereof including oligonucleotides,
ESTs, STSs, etc. Target polynucleotides which may be analyzed by
the methods and compositions of the invention also include RNA
molecules such as, but by no means limited to messenger RNA (mRNA)
molecules, ribosomal RNA (rRNA) molecules, cRNA molecules (i.e.,
RNA molecules prepared from cDNA molecules that are transcribed in
vivo) and fragments thereof.
[0056] The target polynucleotides may be from any source. For
example, the target polynucleotide molecules may be naturally
occurring nucleic acid molecules such as genomic or extragenomic
DNA molecules isolated from an organism, or RNA molecules, such as
mRNA molecules, isolated from an organism. Alternatively, the
polynucleotide molecules may be synthesized, including, e.g.,
nucleic acid molecules synthesized enzymatically in vivo or in
vitro, such as cDNA molecules, or polynucleotide molecules
synthesized by PCR, RNA molecules synthesized by in vitro
transcription, etc. The sample of target polynucleotides can
comprise, e.g., molecules of DNA, RNA, or copolymers of DNA and
RNA. In preferred embodiments, the target polynucleotides of the
invention will correspond to particular genes or to particular gene
transcripts (e.g., to particular mRNA sequences expressed in cells
or to particular cDNA sequences derived from such mRNA sequences).
However, in many embodiments, particularly those embodiments
wherein the polynucleotide molecules are derived from mammalian
cells, the target polynucleotides may correspond to particular
fragments of a gene transcript. For example, the target
polynucleotides may correspond to different exons of the same gene,
e.g., so that different splice variants of that gene may be
detected and/or analyzed.
[0057] In preferred embodiments, the target polynucleotides to be
analyzed are prepared in vitro from nucleic acids extracted from
cells. For example, in one embodiment, RNA is extracted from cells
(e.g., total cellular RNA) and messenger RNA is purified from the
total extracted RNA. cDNA is then synthesized from the purified
mRNA using, e.g., oligo-dT or random primers. In another preferred
embodiment, the target polynucleotides are cRNA prepared from
purified messenger RNA extracted from cells (see, e.g., U.S. Pat.
Nos. 5,891,636, 5,716,785 and 5,545,522; see also, U.S. Pat. No.
6,271,002). Preferably, the target polynucleotides are short and/or
fragmented polynucleotide molecules which are representative of the
original nucleic acid population of the cell.
[0058] The target polynucleotides to be analyzed by the methods and
compositions of the invention are preferably detectably labeled.
For example, cDNA can be labeled directly, e.g., with nucleotide
analogs, or indirectly, e.g., by making a second, labeled cDNA
strand using the first strand as a template. Alternatively, the
double-stranded cDNA can be transcribed into cRNA and labeled.
[0059] Preferably, the detectable label is a fluorescent label,
e.g., by incorporation of nucleotide analogs. Other labels suitable
for use in the present invention include, but are not limited to,
biotin, imminobiotin, antigens, cofactors, dinitrophenol, lipoic
acid, olefinic compounds, detectable polypeptides, electron rich
molecules, enzymes capable of generating a detectable signal by
action upon a substrate, and radioactive isotopes. Preferred
radioactive isotopes include .sup.32P, .sup.35S, .sup.14C, .sup.15N
and .sup.125I. Fluroescent molecules suitable for the present
invention include, but are not limited to, fluorescein and its
derivatives, rhodamine and its derivatives, texas red,
5'carboxy-fluorescein ("FMA"),
2',7'-dimethoxy-4',5'-dichloro-6-carboxy-fluorescein ("JOE"),
N,N,N',N'-tetramethyl-6-carboxy-rhodamine ("TAMRA"),
6'carobxy-X-rhodamine ("ROX"), HEX, TET, IRD40, and IRD41.
Fluroescent molecules that are suitable for the invention further
include: cyamine dyes, including by not limited to Cy3, Cy3.5 and
Cy5; BODIPY dyes including but not limited to BODIPY-FL, BODIPY-TR,
BODIPY-TMR, BODIPY-630/650, and BODIPY-650/670; and ALEXA dyes,
including but not limited to ALEXA-488, ALEXA-532, ALEXA-546,
ALEXA-568, and ALEXA-594; as well as other fluorescent dyes which
will be known to those who are skilled in the art. Electron rich
indicator molecules suitable for the present invention include, but
are not limited to, ferritin, hemocyanin, and colloidal gold.
Alternatively, in less preferred embodiments the target
polynucleotides may be labeled by specifically complexing a first
group to the polynucleotide. A second group, covalently linked to
an indicator molecules and which has an affinity for the first
group, can be used to indirectly detect the target polynucleotide.
In such an embodiment, compounds suitable for use as a first group
include, but are not limited to, biotin and iminobiotin. Compounds
suitable for use as a second group include, but are not limited to,
avidin and streptavidin.
[0060] The target polynucleotides which are analyzed (e.g.,
detected) by the methods and compositions of the invention are
contacted to a probe or to a plurality of probes under conditions
such that polynucleotide molecules having sequences complementary
to the probe hybridize thereto. As used herein, a "probe" refers to
polynucleotide molecules of a particular sequence and to which
target polynucleotide molecules having a particular polynucleotide
sequence (generally a sequence complementary to the probe sequence)
are capable of hybridizing such that hybridization of the target
polynucleotide molecules to the probe can be detected. The
polynucleotide sequences of the probes may be, e.g., DNA sequences,
RNA sequences, or sequences of a copolymer of DNA and RNA. For
example the polynucleotide sequence of the probes may be full or
partial sequences of genomic DNA, mRNA sequences extracted from
cells, cDNA sequences reverse transcribed from RNA (e.g., mRNA)
sequences, or cRNA sequences transcribed from cDNA sequences. The
polynucleotide sequences of the probes may also be synthesized,
e.g., by oligonucleotide synthesis. The probe sequences can also be
synthesized enzymatically in vivo, enzymatically in vitro (e.g., by
PCR), or non-enzymatically in vitro.
[0061] Preferably, the probes used in the methods of the present
invention are immobilized to a solid support or surface such that
polynucleotide sequences which are not hybridized or bound to the
probe or probes may be washed off and removed without removing the
probe or probes and any polynucleotide sequence bound or hybridized
thereto. For example, the probes may comprise double-stranded DNA
comprising genes or gene fragments or sequences derived therefrom
bound to a solid support or surface such as a glass surface or a
blotting membrane (e.g., a nylon or nitrocellulose membrane). In
one particular embodiment, the probes will comprise an array of
distinct oligonucleotide sequences bound to a solid support or
surface, such as a glass surface. Preferably, the array of
sequences is an addressable array. Specifically, each particular
probe (or rather each particular probe sequence) is preferably
located at a particular, known location on the surface or
support.
[0062] Generally, the oligonucleotide sequences will be between 15
and 500 nucleotide bases in length, more preferably between 20 and
100 nucleotide bases in length. However, larger oligonucleotide
sequences (i.e., between 40 and 80 bases in length) are
particularly preferred. Thus, for example, in certain preferred
embodiments the oligonucleotide probe sequences can be between
40-80, 45-80, 50-80, 55-80 bases in length or, alternatively,
between 40-75, 40-70, 40-65, 40-60, 40-55, 40-50, 45-75, 45-70,
45-60, 45-55, 50-75, 50-70, 50-65, 50-60, 55-75, 55-70, 55-65 and
55-60 bases in length. Specific, exemplary oligonucleotide sequence
lengths which may be used as probes in the present invention
include oligonucleotide sequences which are 20, 25, 30, 35, 40, 45,
50, 55 and 60 bases in length. Sequences of about 50 to 60 bases in
length are particularly preferred.
[0063] Longer oligonucleotide sequences can be readily identified
which hybridize both more specifically and more sensitively to a
particular target polynucleotide sequence than do shorter
oligonucleotide sequences (e.g., less than 40 bases in length) and
longer full length DNA sequences (e.g., full length cDNA
sequences). FIG. 13, which plots the specificity of oligonucleotide
probes of various lengths and under various levels of hybridization
stringency, demonstrates this point by way of example.
Specifically, oligonucleotide microarrays were synthesized,
according to the methods described in Section 5.3 below, which
comprised an overlapping series of different length
oligonucleotides from the S. cerevisiae gene HXT3 (Ko et al., 1993,
Mol. Cell. Biol. 13:638-648; GenBank Accession No. L07080). The
oligonucleotides included sequences 20, 25, 30, 35, 40, 45, 50, 55
or 60 bases in length and beginning at every third position in the
sequence so that oligonucleotide sequences of each specific length
"tile" through the complete HXT3 gene sequence (e.g., 20mers that
spanned positions 1-20, 4-23, 7-26, etc.). The microarrays were
simultaneously hybridized with a Cy3-labeled cRNA sample from an
strain of S. cerevisiae bearing a homozygous deletion in the HXT3
gene (i.e., an HXT3 gene non-specific sample) and with a
Cy5-labeled cRNA sample corresponding the HXT3 gene sequence (i.e.,
an HXT3 gene specific sample) in the presence of increasing
concentrations of formamide (16%, 32% and 48%) which correlate with
hybridization stringency. The absolute hybridization intensity and
the ratio of hybridization intensity of the HXT3 gene specific to
the non-specific sample were determined, as described, e.g., in
Section 5.2 below and in FIG. 6. In particular, the ratio of the
HXT3 gene specific to non-specific hybridization, which is a
measure of hybridization specificity, is plotted in FIG. 13 for the
different probe lengths and hybridization intensities. As can be
seen from the figure, longer oligonucleotide probes (e.g., greater
than about 40 bases in length and more preferably 55-60 bases in
length) are significantly more specific for the target
polynucleotide sequence, particularly under hybridization
conditions of higher stringency (e.g., higher levels of
formamide).
[0064] The invention is based, at least in part, on the discovery
that the number of probe sequences required to reliably and
accurately report a particular polynucleotide sequence, such as the
sequence of a particular gene, may be reduced to as few as one
oligonucleotide probe by carefully selecting probes according to
the methods described herein. Thus a user can both efficiently and
accurately detect, e.g., expression levels of a large number of
genes and/or gene products by minimizing the number of probes
required to detect each gene or gene transcript according to the
methods described herein. For example, using the ranking and/or
selection methods of the invention, a user can select specific
probes, e.g., for "screening chips" that may be used to screen for
expression levels a substantial portion, or even all of the genes
or gene transcripts of a particular organism. The invention
therefore also provides such screening chips as well as methods for
obtaining such screening chips. In certain preferred embodiments,
such screening chips will have probes specific to least about 50%
of the genes in the genome of an organism, more preferably to at
least about 75%, still more preferably to at least about 85%, even
more preferably to at least about 90%, and still more preferably to
at least about 99%. In fact most preferably, the screening chips of
the invention have probes specific to all of the genes (i.e., 100%)
in the genome of an organism. In other embodiments, however, the
screening chips have probes specific for those particular genes
expressed by a particular cell or cell type of interest. In such
embodiments, a screening chip will therefore preferably have probes
specific for all of the genes expressed by the cell or cell type of
interest, which will often be substantially less than 50% of the
genes in the entire genome of the cell or organism (e.g., 20%).
[0065] The organism may be of any species, including procaryotic
organisms, such as E. coli and other bacteria, and eukaryotic
organisms including, but not limited to, Saccharomyces cerevisiae.
The organism may also be a higher, multi-cellular organism such as
a plant or animal, including mammalian animals such as humans.
Preferably the screening chips of the invention comprise no more
than 10, more preferably no more than 5 and most preferably only
one probe for each target polynucleotide.
[0066] In preferred embodiments, such "screening chips" may be used
to identify "signature genes," i.e., those genes or gene
transcripts that are of particular interest to a user. For example,
signature genes can comprise those genes or gene transcripts that
are most responsive to a particular perturbation or to a particular
class of perturbations. Exemplary types and classes of
perturbations including, exposure to one or more drugs, including
drugs from a particular family of drugs. Other exemplary types and
classes of perturbations can include viral infection, including
infection by a particular type or family of virus, or certain types
or classes of disease, such as cancer and immune disorders, to name
a few.
[0067] The ranking and/or selection methods of the invention may be
used to select particular probes for efficiently and accurately
detecting changes in the expression levels of those signature
genes, e.g., on a "signature chip." Thus, such signature chips are
also provided by the present invention. In specific embodiments, a
signature chip of the invention can contain probes specific for as
many as 2,000 or more of the genes most responsive to a particular
change or perturbation to the cell or organism. More preferably,
however, signature chips comprise probes specific for the 5, 20,
50, 100 or more of the genes most responsive to a particular change
or perturbation to the cell or organism. In certain preferred
embodiments, the signature chips comprise at least two probes
(i.e., one probe pair) for each target polynucleotide: a "match
sequence" probe, which is complementary to a particular target
polynucleotide; and at least one "mismatch sequence" probe, whose
polynucleotide sequence is only partially complementary (e.g.,
contains 1-3 mismatched bases) to the particular target
polynucleotide. In such embodiments, the signature chips preferably
comprise at least five match sequence probes for each target
polynucleotide, more preferably at least 10 match sequence probes,
still more preferably at least 20 match sequence probes. In such
embodiments, the signature chips comprise at least 1 mismatch
sequence probe for each match sequence probe. More preferably,
however, the signature chips comprise a plurality of mismatch
sequence probes (e.g., 4 to 10) for each match sequence probe.
[0068] In other, more preferred embodiments, the signature chips
comprise only match sequence probes for each target polynucleotide.
In such embodiments, the signature chips preferably comprise at
least 5, more preferably at least 10, and still more preferably at
least 15 (e.g., between 15-20), match sequence probes for each
target polynucleotide.
[0069] The methods and compositions of the invention are described,
in detail, below. In particular, Section 5.1 provides a general
overview of the ranking and selection methods of the invention, as
well as the screening chips and signature chips which are designed
according to such methods. Section 5.2 describes, in detail, the
preferred analytical systems used to practice the methods described
in Section 5.1. Section 5.3 provides exemplary systems, such as
microarrays, which can be used to measure hybridization and/or
cross hybridization levels and which can therefore be used in the
methods of the present invention.
[0070] The detailed description is by way of several exemplary
illustrations, in increasing detail and specificity, of the general
methods and compositions of the invention. These examples are
non-limiting and related variants that will be apparent to one
skilled in the art are intended to be encompassed by the appended
claims.
5.1. OVERVIEW OF THE INVENTION
[0071] A flow chart illustrating an exemplary, non-limiting
embodiment of the general methods of the present invention is shown
in FIG. 1. In this particular embodiment, hybridization conditions
are first provided or determined (101), as described in subsection
5.1.1 below, to optimize the specificity of each probe for its
target polynucleotide sequence. Candidate oligonucleotide probes
are then ranked and/or selected (102), according to the methods
described below in subsection 5.1.2, based on their sensitivity and
specificity for their target sequences, and screening chips are
then synthesized (103) using the selected probes. In preferred
embodiments, the candidate probes are further ranked and selected
according to empirical, iterative methods (FIG. 1; steps 104-105)
which are described below in subsection 5.1.3. For example, in
preferred embodiments, no more than 10, no more than 5, no more
than 4, no more than 3, or no more than 2 candidate probes are
identified for each target polynucleotide. Most preferably, one
candidate probe is identified for each target polynucleotide. The
ranking and/or selection methods described herein are particularly
useful to design both screening chips and signature chips which may
be used, e.g., to examine changes in genetic expression in cells.
Briefly, the screening chips of the invention are particularly
useful for applications wherein a sample must be screened for a
large number of polynucleotide sequences, e.g., the entire genome
of an organism or a substantial fraction thereof. By contrast the
signature chips of the invention are most useful for obtaining an
accurate measurement of changes in the level of a relatively small
number of polynucleotide sequences in a sample, such as changes in
the expression of certain specific genes of interest to a user.
Such screening and signature chips are therefore also considered a
part of the present invention and are described below in Sections
5.1.4 and 5.1.5, respectively.
5.1.1. HYBRIDIZATION CONDITIONS
[0072] Hybridization conditions, such as conditions of salt and
temperature, that are appropriate for hybridizing target
polynucleotide molecules to one or more probe sequences are
generally well known in the art. For example, conditions of higher
temperature and lower salt concentration, or "high stringency," are
generally preferred to minimize cross-hybridization. Exemplary
highly stringent conditions comprise hybridization to filter-bound
DNA in 5.times.SSC, 1% sodium dodecyl sulfate (SDS), 1 mM EDTA at
65.degree. C. followed by post hybridization washing in
0.1.times.SSC/0.1% SDS at 68.degree. C. (Ausubel et al., Eds.,
1989, Current Protocols in Molecular Biology, Vol. I, Green
Publishing Associates, Inc., and John Wiley & Sons, Inc., New
York at p. 2.10.3). Conditions of high stringency can also be
produced by addition of a denaturant such as formamide.
Particularly preferred hybridization conditions comprise:
incubation for 12-24 hours at, e.g., 40.degree. C., in 1 M NaCl, 50
mM MES buffer (pH 6.5), 0.5% sodium sarcosine and 30%
formamide.
[0073] In particular, the hybridization conditions used in the
methods of the invention are preferably such that the amount of
specific hybridization is maximized while the amount of
cross-hybridization or non-specific hybridization is minimized. In
those preferred embodiments where target polynucleotides hybridize
to oligonucleotide probes, specificity may be maximized by
hybridizing at a temperature that is at or near (e.g., within
2.degree. C. or within 5.degree. C.) the melting temperature
("T.sub.m") of the target polynucleotide and probe. This fact is
illustrated, e.g., in FIGS. 2 and 3. FIG. 2 shows an exemplary,
calculated "melting curve" for a perfect match and two imperfect
match duplexes. Specifically, FIG. 2 depicts a plot of the
predicted fraction of target polynucleotides bound to an
oligonucleotide probe as a function of the hybridization
temperature. The "melting temperature" of any given target
polynucleotide to the probe is defined in the art to mean the
temperature at which exactly one-half (i.e., 50%) of the target
polynucleotide molecules in a sample are bound to the probe. Thus,
the melting temperature is the point on the melting curve at which
the bound fraction of polynucleotide molecules is 0.5 (e.g.,
58.degree. C. for the perfect-match duplex in FIG. 2)
[0074] The plot depicted in FIG. 2 shows, not only the predicted
fraction of perfect match target polynucleotide molecules bound to
the probe (PM), but also the fraction of bound polynucleotides
having one base mismatch (1 MM) or two base mismatches (2MM) to the
probe. The plot depicted in FIG. 2 also shows the ratio of bound
perfect match to bound 1 base or 2 base mismatch target
polynucleotides at a given temperature. By maximizing these ratios,
the amount of specific (i.e., PM) hybridization is maximized while
minimizing cross hybridization, e.g., from 1MM and 2MM in FIG. 2.
Inspection of FIG. 2 reveals that these ratios are maximized at
about 62.degree. C., i.e., slightly above the melting temperature
of the perfect match duplex.
[0075] FIG. 3 shows the experimental verification of this
principle. Specifically, oligonucleotide probes of either 22 or 35
nucleotides in length (i.e., 22-mer's or 35-mers) were synthesized
using standard inkjet printing techniques known in the art (see
Blanchard, International Patent Publication WO 98/41531, published
Sep. 24, 1998; Blanchard et al., 1996, Biosensors and
Bioelectronics 11:687-690; Blanchard, 1998, in Synthetic DNA Arrays
in Genetic Engineering, Vol. 20, J. K. Setlow, Ed., Plenum Press,
New York at pages 111 - 123). Each of the synthesized probes was
either a perfect match, single or double base mismatch to a
specific target polynucleotide sequence. FIG. 3A shows a plot of
the observed ratio of hybridization intensities to this target
polynucleotide sequence between the perfect match and single
mismatch 22-mer probes (PM/SM, solid line) and between the perfect
match and double mismatch 22-mer probes (PM/DM, dashed line). A
histogram showing the predicted perfect-match melting temperature
for each 22-mer probe is shown in FIG. 3B. FIGS. 3C and 3D show an
identical analysis of the hybridization and predicted melting
temperatures of the 35-mer probes. In both FIGS. 3A and 3C, the
highest ratio (and hence highest specificity) between perfect match
and mismatch probes is obtained at a hybridization temperature that
is equal to or slightly above the median melting temperature in
FIGS. 3B and 3D, respectively.
[0076] Methods for determining the melting temperature of a
particular polynucleotide duplex are well known in the art and
include, e.g., predicting the melting temperature using well known
physical models adapted to experimental data (see, e.g.,
SantaLucia, J., 1998, Proc. Natl. Acad. Sci. U.S.A. 95:11460-1465
and the references cited therein). Mathematical algorithms and
software for predicting melting temperatures using such models are
readily available as described, e.g., by Hyndman et al., 1996,
Biotechniques 20:1090-1096. For example, the melting temperature
for an RNA/DNA duplex 25 base pairs in length in 1 M salt solution
is between about 60 to about 70.degree. C.
[0077] In preferred embodiments, the methods of the invention are
practiced using a plurality of oligonucleotide probes, e.g., in a
microarray such as those described in Section 5.3 below. In such
embodiments, it is generally not feasible or desirable to select
and/or use individual hybridization conditions that are optimized
for each individual probe. Rather, a single set of hybridization
conditions is preferably selected and used that optimizes
hybridization of polynucleotide molecules overall to all of the
oligonucleotide probes. For example, in such embodiments the
melting temperatures of the perfect match polynucleotide molecules
from each probe will typically fall within some range of
temperatures. In such embodiments, therefore, the hybridization
temperature is selected to be near or at the upper limit of this
range.
5.1.2. PROPERTIES AFFECTING TARGET AND CROSS-HYBRIDIZATION
[0078] Candidate oligonucleotide probes are ranked and/or selected
based on at least two, and preferably on a plurality of properties
and/or parameters. For example, in the exemplary embodiment
illustrated in FIG. 1, candidate oligonucleotide probes are ranked
and/or selected (102) based on both their sensitivity and
specificity for their target polynucleotide sequence. As used
herein, the "sensitivity" of a probe refers to the fraction of
molecules of the probe that hybridize to polynucleotide molecules
(or that have polynucleotide molecules hybridized thereto) under a
particular set of hybridization conditions (e.g., the selected or
provided hybridization conditions). The "specificity" of a probe,
as used herein, is understood to refer to the ratio of target
(e.g., perfect match) polynucleotide molecules to non-target
polynucleotide molecules hybridized to the probe under a particular
set of hybridization conditions (e.g., the selected or provided
hybridization conditions).
[0079] Other properties and parameters by which candidate
oligonucleotide probes can be ranked including, but are not limited
to: (a) probe size or length; (b) binding energies, including both
the perfect match duplex (i.e., of a probe and its target,
complementary nucleotide sequence) and cross-hybridization binding
energies; (c) base composition, including, for example, the
relative amount or percentage of one or more particular nucleotide
bases (e.g., adenine, guanine, thymine or cytosine) in a probe
sequence, as well as the relative amount or percentage of any
combination of such nucleotide bases; (d) the position of a probe's
complementary sequence in the sequence of its "target"
polynucleotide or gene sequence; and (e) probe sequence complexity,
including the presence or lack of common repetitive elements such
as polynucleotide repeats (i.e., simple, contiguous repeats of one
or more nucleotide bases) as well as more complicated repetitive
elements that are well known in the art. Still other exemplary
parameters which can be used in the methods and compositions of the
invention for ranking and/or selecting oligonucleotide probes
include: (f) self dimer binding energy (i.e., the tendency for a
particular probe to hybridize to its own sequence); (g) the
structure content of the complementary, target polynucleotide
sequence for a particular probe (e.g., the presence or absence of
certain structural features or motifs); and (h) the information
content of a probe's nucleotide sequence. Each of these properties
is discussed, in detail, hereinbelow.
[0080] Preferably, the target polynucleotide sequence of a
candidate probe is its "perfect match" sequence, i.e., to
polynucleotide molecules that comprise the nucleotide sequence that
is complementary to the sequence of the oligonucleotide probe and
which, therefore, hybridize to the probe with no mismatches.
Evaluating Binding Energy:
[0081] Both the sensitivity and specificity of a particular probe
depends upon the binding energies, .DELTA.G, of polynucleotide
molecules to the probe, as shown in FIG. 4. In particular, the
hybridization intensity traces a sigmoidal curve which follows the
melting curve of the probe, decreasing as the binding energy
increases. Specificity, however, is maximum at or slightly above
the melting temperature of the probe (i.e., at a binding energy
that is equal to or slightly greater than zero). Thus, in preferred
embodiments, the sensitivity and/or specificity are determined or
predicted from the binding energies.
[0082] It is noted that the skilled artisan readily appreciates
that the term "binding energy," as used herein, refers to the
difference of the energy of polynucleotide molecules (e.g., a
target polynucleotide and a polynucleotide probe) when they are in
a bound state (i.e., when they are bound or hybridized to each
other) from when they are in an unbound state. This definition is
readily expressed mathematically by the formula
.DELTA.G=G.sub.bound-G.sub.unbound. Thus, a polynucleotide that has
the "largest" binding energy to a particular probe is one for which
the difference between the energies of the bound and unbound
polynucleotide is greatest. In particular and as the skilled
artisan also readily appreciates, because the energy of
polynucleotides in a bound state is ordinarily lower than the
energy of the unbound polynucleotides, the binding energy (i.e.,
.DELTA.G) will ordinarily be a negative number. Thus, as used
herein, the polynucleotide having the "largest" binding energy to a
particular probe will, in fact, be the polynucleotide for which
.DELTA.G is the most negative.
[0083] Binding energies for polynucleotide duplexes, and
particularly for oligonucleotide duplexes in solution, may be
readily obtained or predicted, at least in part, by using
theoretical models known in the art, including, e.g.,
"nearest-neighbor" models such as those described by SantaLucia,
1998, Proc. Natl. Acad. Sci. US.A. 95:1460-1465. Such models assume
that the stability (i.e., the binding energy) of individual
base-pairs in a polynucleotide duplex depends upon the identity and
orientation of the neighboring base pairs. The binding energy
.DELTA.G is therefore expressed as a sum of the free energies of
the individual dimer duplexes and "initiation factors" for duplex
formation. Thus, for example, for DNA/RNA complexes there are 16
unique Watson-Crick dimer duplexes which are listed in Table I,
below. TABLE-US-00001 TABLE I DNA/RNA WATSON-CRICK DIMER DUPLEXES
5'-A A-3' 5'-T A-3' 5'-C A-3' 5'-G A-3' 3'-U U-5' 3'-A U-5' 3'-G
U-5' 3'-C U-5' 5'-A T-3' 5'-T T-3' 5'-C T-3' 5'-G T-3' 3'-U A-5'
3'-A A-5' 3'-G A-5' 3'-C A-5' 5'-A C-3' 5'-T C-3' 5'-C C-3' 5'-G
C-3' 3'-U G-5' 3'-A G-5' 3'-G G-5' 3'-C G-5' 5'-A G-3' 5'-T G-3'
5'-C G-3' 5'-G G-3' 3'-U C-5' 3'-A C-5' 3'-G C-5' 3'-C C-5'
[0084] Thus, for example, the binding energy of a particular 16-mer
polynucleotide duplex may be determined, according to a nearest
neighbor model, by the equation: .DELTA. .times. .times. G = i = 1
16 .times. q i .times. .DELTA. .times. .times. g + q 17 .times.
.DELTA. .times. .times. g 17 + q 18 .times. .DELTA. .times. .times.
g 18 ( Eq . .times. 1 ) ##EQU1## In particular, in Equation 1
above, .DELTA.g.sub.i is the binding energy of the i.times.th
individual dimer duplex and qi is the number of occurrences of the
i'th individual dimer duplex in the polynucleotide complex of
interest. Two "initiation parameters" .DELTA.g.sub.17 and
.DELTA.g.sub.18 are also used in the model, and q.sub.17 and
q.sub.18 are the number of terminal (i.e., end) base pairs that are
A/T and G/C base pairs, respectively. Generally, the individual
dimer-duplex binding energies and initiation parameters have values
that are already known in the art (see, e.g., SantaLucia, supra).
Alternatively, such parameters can be experimentally determined,
e.g., by a user, as explained below.
[0085] It is understood that the nearest neighbor models used in
the methods of the present invention may comprise additional
binding energy terms besides the initiation parameters and dimer
binding energies discussed above. In particular, the nearest
neighbor models of the invention can also be used to calculate or
predict the binding energy of polynucleotide duplexes comprising
one or more mismatched base pairs by including binding energy terms
for additional dimer-duplexes that contain a base-pair mismatch.
Dimer-duplex binding energies for such mismatch dimer-duplexes can
be obtained or determined according to the same methods as those
used to obtain or determine the binding energy terms for
Watson-Crick dimer duplexes, including the methods described
hereinbelow.
[0086] In more preferred embodiments of the invention wherein
surface bound polynucleotides are employed, such as in microarrays,
binding energies can-also be estimated from experiments using the
same surface-bound polynucleotide probes. Thus, in particularly
preferred embodiments wherein the methods of the invention are used
to design microarray "chips," the binding energies for
oligonucleotide probes can be estimated using the same chips that
are being designed. For example, in certain particular, but
non-limiting, embodiments, the binding energies of a set of
oligonucleotide probes, p=1 to N, on a microarray of N probes can
be determined by "spiking" polynucleotide molecules corresponding
to a particular sequence into the hybridization solution.
Specifically, a concentration, c, of polynucleotide molecules
corresponding to the particular sequence is hybridized to the
probes of the microarray, and the hybridization level, I.sub.p, of
the particular polynucleotide molecules to each probe p is
measured. Under the preferred hybridization conditions (i.e., high
stringency) of the present invention, the hybridization level is
related to the polynucleotide concentration and the binding energy,
.DELTA.G.sub.p, according to Equation 2:
I.sub.p=sM(c.sub.p)e.sup..DELTA.G.sub.P.sup./RT (Eq. 2) wherein R
denotes the ideal gas constant (1.9872 kcal/mol.degree. K), and T
is the hybridization temperature (in degrees Kelvin). s denotes a
correction factor, e.g., for detector and label incorporation
efficiencies, and M(c.sub.p) is a function related to the
concentration of the polynucleotide molecules corresponding to the
known sequence. Pursuant to Equation 2 above, the log(intensity) of
hybridization is linearly related to the binding energy, i.e.,: log
.function. ( I p ) = log .function. [ s M .function. ( c p ) ] +
.DELTA. .times. .times. G p RT ( Eq . .times. 3 ) ##EQU2##
[0087] In one embodiment, therefore, the hybridization level may be
measured or determined for a fixed concentration, c.sub.p, of
polynucleotide molecules at a plurality of hybridization
temperatures T. The binding energy .DELTA.G.sub.p can then be
determined from the slope of the line log(I.sub.p) v. 1/T. More
preferably, however, Equation 3 can also be used to determine the
parameters .DELTA.g.sub.i for use in a nearest neighbor model
(i.e., in Equation 1, above). In particular, by using the
expression for .DELTA.G provided in Equation 1, Equation 3 can also
be expressed as: log .function. ( I p ) = log .function. [ s M
.function. ( c ) ] + i = 1 18 .times. q i , p RT .times. .DELTA.
.times. .times. g i ( Eq . .times. 4 ) ##EQU3## wherein q.sub.p,i
denotes the number of occurrences of the i'th individual dimer
duplex in the polynucleotide complex of probe p.
[0088] As one skilled in the art readily appreciates, the binding
energies of individual dimer duplexes and the initiation parameters
(i.e., the .DELTA.g.sub.i) may be determined from Equation 4 above
using techniques of mathematical analysis known in the art. For
example, Equation 4 may also be represented as an equation of
vectors and matrices: log(I)=log [sM(c)]+Q.DELTA.g (Eq. 5)
Specifically, I in Equation 5 denotes the vector of hybridization
intensities {I.sub.p} for each probe p of the microarray. .DELTA.g
denotes the "dimer binding energy vector," i.e., the vector of
binding energies of individual dimer duplex and initiation
parameters (i.e., {.DELTA.g.sub.i}), and Q is the matrix of
elements {q.sub.p,i}. Thus, provided hybridization intensities for
individual probes of a microarray, the dimer binding energy vector
may be readily determined, e.g., from a least-squares solution:
.DELTA..sub.g LSQ=(Q.sup.TQ).sup.-1Q.sup.T log(I) (Eq. 6) in which
Q.sup.T denotes the transpose of the matrix Q. Alternatively, a
conditioned least squares solution may be used as provided by the
equation: .DELTA.g.sub.LSQ=(Q.sup.TQ+.LAMBDA.).sup.-1Q.sup.T log(I)
(Eq. 7) wherein A is a scaled version of the identity matrix which
is optionally used, e.g., to keep the sizes of the elements of
.DELTA.g to within limits, e.g., determined or provided. For
example, generally a user will prefer to keep the elements of
.DELTA.g less than or equal to about 8 kcal/M in magnitude. As will
be appreciated by one skilled in the art, it is understood that in
Equations 6 and 7 above, the constant term sM(c.sub.p) is subsumed
into the definition of .DELTA.g. Numerical techniques for solving
linear equations such as Equations 6 and 7 above are well known in
the art and include, e.g., the numerical methods and algorithms
described by Press et al. (1992, Numerical Recipes in C, Chapter 2:
"Solution of Linear Algebraic Equations," Cambridge University
Press).
[0089] Although the expression cannot be readily expressed in a
linear form as in Equation 3 above, under less preferred
hybridization conditions (e.g., of low stringency or moderately
stringent hybridization conditions), as one skilled in the art
readily appreciates, binding energies and binder parameters (i.e.,
the .DELTA.G and .DELTA.g terms in Equations 3 and 4 above) can
nevertheless be obtained or determined from similar systems of
equations using methods of analytical and numerical analysis known
in the art.
Prediction of Probe Sensitivity and Specificity:
[0090] Once binding energies for polynucleotide molecules to the
probes are provided or determined, e.g., using a nearest-neighbor
model with appropriate parameters .DELTA.g.sub.i, both the
sensitivity and specificity of a probe can be readily predicted,
e.g., using theoretical models. As discussed above, the level of
hybridization of a particular polynucleotide sequence p to a given
probe is directly related to the binding energy .DELTA.G.sub.p of
that sequence to the probe. More specifically, the level of
"target" hybridization T.sub.p, i.e., the level of hybridization of
a target polynucleotide sequence p to a particular probe is
specified by: T p = s M .function. ( c p ) .times. k .times. exp
.function. ( .DELTA. .times. .times. G kp RT ) ( Eq . .times. 8 )
##EQU4## wherein .DELTA.G.sub.k,p is the binding energy of the
duplex between the probe and the sequence starting at position k on
the target polynucleotide. M(c.sub.p) is the concentration (i.e.,
abundance) of the target polynucleotide in the hybridization
sample, and s denotes a correction factor as explained supra.
Likewise the level of cross-hybridization of all other "non-target"
polynucleotide sequences j.noteq.p is specified by: X p = s j
.noteq. p .times. M .function. ( c j ) [ k .times. exp .function. (
.DELTA. .times. .times. G kj RT ) ] ( Eq . .times. 9 ) ##EQU5##
wherein .DELTA.G.sub.kj is the binding energy of the probe starting
at position k on the non-target polynucleotidej, and M(c.sub.j) is
the concentration (i.e., abundance) of the polynucleotidej in the
hybridization sample. Thus, the specificity of the probe for the
sequence p is provided by: S p = T p X p ( Eq . .times. 10 )
##EQU6## As will be readily appreciated by one skilled in the art,
it is understood that the specificity provided by Sp in Equation 10
above is independent of the value of the correction factor s in
Equations 8 and 9.
[0091] In many embodiments, the actual abundance M(c.sub.j) of at
least some polynucleotide sequences in a sample will not be known.
In such embodiments, it is preferable to set the value of the
abundances to unity when evaluating Equations 8-10, above. Also, in
many embodiments, the methods and compositions of the invention are
used to evaluate polynucleotide expression in cells, such as
mammalian cells, whose genomes contain repetitive sequences (see,
e.g., Claverie, 1996, Methods in Enzymology 266:212-227). In such
embodiments, it is preferable to eliminate candidate probes
corresponding to such repetitive sequences before evaluating
candidate probe sensitivities and specificities. It is still
further preferable to eliminate candidate probes corresponding to
other sequences of low information content, as explained
hereinbelow, before evaluating the sensitivity and specificity of
candidate probes.
[0092] In most preferred embodiments of the invention, i.e.,
wherein the methods and compositions of the invention are used to
evaluate the entire genome of an organism, complete evaluation of
Equation 8-10 requires the evaluation of more than 10.sup.14
exponential terms. Therefore, it is preferable to make certain
approximations before evaluating these equations so that the number
of numerical calculations is reduced to a manageable size. For
example, in certain embodiments, probe candidates are first
selected on the basis of their determined or predicted binding
energy .DELTA.G, so that only probes having or predicted to have at
least a certain minimum binding energy or within some interval of
binding energies (determined, e.g., by a user) are evaluated. Still
more preferably, before evaluating candidate probes for
specificity, the probes are ranked and/or selected according to one
or more of the properties described below, such as size and length,
base composition, sequence complexity and/or combinations thereof.
In other embodiments, wherein the relative abundances of at least
some polynucleotide species in a sample are known, only the most
abundant polynucleotide sequences are considered when evaluating
specificity, e.g., using Equation 10 above. For example, in certain
embodiments, only those polynucleotide sequences in a sample which
represent, in toto, at least 50% of the total number of
polynucleotide molecules in the sample are considered.
Alternatively, only those polynucleotide sequences in a sample
which represent, in toto, at least 75%, 80%, 85%, 90%, 95%, or 99%
of the total number of polynucleotide molecules in the sample are
considered.
[0093] In particularly preferred embodiments, a homology search
method such as BLAST ("Basic Local Alignment Search Tool") and
PowerBLAST (see, in particular, Altschul et al., 1990, J. Mol.
Biol. 215:403-410; Altschul, 1997, Nucleic Acids Res. 25:3389-3402;
and Zhang and Madden, 1997, Genome Res. 7:649-656) are first
performed against each probe sequence to identify polynucleotides,
e.g., in a database of expressed sequences such as the GenBank or
the dbEST database, which comprises sequences that are most
identical or homologous to each probe's complementary sequence. For
example, in preferred embodiments, sequences which are at least
50%, 60%, 70%, 80%, 90%, 95%, or 99% identical to a probe's target
sequence are identified using a search algorithm such as BLAST or
PowerBLAST according to its default parameters. Preferably the
search algorithm is employed using parameters set to detect
perfect-match sequences of a seed length of, e.g., 7 to 15 or, more
preferably, 7 to 12 bases. Binding energies and binding specificity
are then evaluated only for polynucleotide sequences identified in
such searches.
[0094] Preferably, the database of sequence used in such a homology
search is a database of or containing all or substantially all of
the polynucleotide sequences that are present or are believed to be
present in a polynucleotide sample that the probe or probes are
intended to assay. Thus, for example, in embodiments wherein the
sample is a polynucleotide sample (e.g., mRNA, cDNA or cRNA)
derived from a cell or organism, the database is a database of or
containing all or substantially all of the polynucleotide sequences
expressed by that cell or organism. Such a database can contain,
for example, sequences corresponding to 50%, 60%, 70%, 80%, 90%,
95%, 99% or 100% of the polynucleotide sequences expressed by the
cell or organism. The database can also contain 50%, 60%, 70%, 80%,
90%, 95%, 99% or 100% of the gene sequences in the genome of the
cell or organism.
[0095] In a particularly preferred aspect of this embodiment,
polynucleotide sequences that are most identical or homologous to
each probe's complementary sequence are identified, e.g., using a
homology search method such as BLAST or PowerBLAST and their
binding (i.e., cross-hybridization) energies to the probe or probes
are evaluated (e.g., using the nearest neighbor model and Equation
1, described above). In this embodiment, the strongest binding
energy calculated, which is referred to herein as the "minimax
score", is used in place of the score provided in Equation 10,
above, as an indication of the probe's predicted
cross-hybridization. Preferably, weighting factors are not used in
determining the minimax score for a particular probe. However, in
embodiments where relative abundances of the polynucleotide
sequences in the sample are known or can be estimated, the
calculated cross-hybridization energy of each homologous sequence
to the probe can be multiplied by a weighting factor that is
proportional to the homologous sequence's actual or estimated
abundance in the sample. The product of the calculated
cross-hybridization energy and the weighting factor is then used to
determine the minimax score. Alternatively, in cases when limited
abundance information is available the abundances of nontarget
polynucleotide sequences under consideration can be classified into
a limited number of abundance categories (e.g., high and low), and
a simplified set of weighting factors can be used for each category
(e.g., 1 and 0 or, alternatively, 10 and 1 for "high" and "low,"
respectively).
[0096] The sensitivity and/or specificity of a particular probe may
also be determined experimentally, e.g., using differentially
labeled polynucleotide samples. For example, FIGS. 5A and 5B show
observed target and non-target hybridization, respectively, to
oligonucleotide (25-mer) probes that match different positions
along the known S. cerevisiae gene YER019W (GenBank Accession No.
U18778). FIGS. 6A and 6B show observed target and non-target
hybridization, respectively, to oligonucleotide probes that match
different positions along the known S. cerevisiae gene HXT3 (Ko, C.
H. et al., 1993, Mol. Cell. Biol. 13:638-648; GenBank Accession No.
L07080). The ratios of target to non-target hybridization for
probes matched to YER019W and HXT3 are shown in FIGS. 5C and 6C,
respectively. Specifically, the data for each figure were obtained
by hybridization of two differently-labeled samples to the same
oligonucleotide array in accordance with the methods described in
copending provisional U.S. Patent Application Ser. No. 60/154,563,
filed Sep. 17, 1999. One sample contained only target sequences
whereas the other sample was derived from a yeast strain in which
the target gene (YER019W or HXT3) was deleted and so represents
actual cross-hybridization from the remainder of the genome.
Probe Size and Length:
[0097] The present inventors have discovered that both the
sensitivity and specificity of oligonucleotide probes increase with
oligonucleotide length (i.e., with the number of nucleotide bases
in the probe). Thus, oligonucleotide probes can also be ranked
and/or selected in the methods and compositions of the present
invention according to their size or length.
[0098] The probes of the present invention are preferably selected
to be at least 15 bases in length, and are more preferably at least
20 bases in length, more preferably at least 30 bases in length,
more preferably at least 40 bases in length, more preferably at
least 50 bases in length or more preferably 60 bases in length.
[0099] Typically, synthetic nucleotide probe sequences (e.g.,
oligonucleotide sequences) are shorter than 500 bases in length,
and are more typically shorter than 100 bases in length.
Preferably, the probe lengths selected are short enough that
synthesis of pure (i.e., sufficiently pure for use as probes)
full-length sequences is practical using existing techniques (such
as N-phosphonate or phosphoramidite chemistry techniques described,
e.g., in Froehler et al., 1986, Nucleic Acid Res. 14:5399-5407; and
in McBride et al., 1983, Tetrahedron Lett. 24:246-248). Thus, in
preferred embodiments, the oligonucleotide probes selected are 100
or fewer bases in length and are more preferably 90 or fewer bases
in length, more preferably 80 or fewer bases in length, more
preferably 70 or fewer bases in length or more preferably 60 or
fewer bases in length. Thus, in a most preferred embodiment, the
probe nucleotide sequences of the present invention are 40-70 or
50-60 nucleotides in length.
Base Composition:
[0100] The methods and compositions of the present invention can
also be used to rank and/or select oligonucleotide probes according
to their base composition. "Base composition", as the term is used
herein, is understood to refer to the amount or number of
nucleotide bases having a particular chemical identity. Thus, for
example, oligonucleotide probes can be ranked and/or selected in
the methods of the present invention on the basis of the percentage
or fraction of bases that are cytosine ("C"), guanine ("G"),
thymine ("T"), adenine ("A") or, in embodiments where RNA probes
are used, uracil ("U").
[0101] Oligonucleotide probes can also be ranked or selected in the
methods of the present invention based on any mathematical
combination of two or more nucleotide identities. For example, and
not by way of limitation, probes can be ranked and/or selected
based on the percentage or fraction of bases that are either
guanine or cytosine ("G+C %") or, alternatively, based on the
percentage or fraction of bases that are either adenine or thymine
("A+T"). In another embodiment, oligonucleotide probes can be
ranked or selected according to a differential between the percent
or fraction of two or more nucleotide identities, such as the
difference between the percent or fraction of bases in a probe that
are adenine and the percent or fraction of bases that are cytosine
("A-C %").
[0102] In preferred embodiments, oligonucleotide probes are ranked
and/or selected to minimize the number of G and C bases. In
particular, it is already well known in the art that
guanine-cytosine base pairs have a higher stability than do
adenine-thymine base pairs and, further, that many guanine
containing mismatches have a higher stability than do non-guanine
containing mismatches (see, e.g., SantaLucia, 1998, Proc. Natl.
Acad. Sci. US.A. 95:1460-1465). As a result, although the
percentage of guanine-cytosine base pairs is therefore somewhat
correlated with the perfect match duplex binding energy discussed
above, high number of guanine-cytosine base pairs are also
correlated with higher levels of cross-hybridization. Thus, probe
sequences with a low ratio of G-C base pairs (i.e., a low G+C %)
are preferred. Preferably the percentage of G-C base pairs is
between 0 and 75%, more preferably between 0 and 55%, and still
more preferably between 8 and 45%.
[0103] For example, FIG. 16 shows a plot of differential expression
(vertical axis) of the gene ETR103 (GenBank Accession No. M62829)
in unactivated and activated human lymphoblast cells as reported by
several candidate oligonucleotide probes. The fraction of guanine
and cytosine bases in the probes is indicated on the horizontal
axis. Although the difference in the level of ETR103 expression in
unactivated and activated human lymphoblasts is known to be very
high, only those probes with a G+C % less than 0.4 (i.e., 40%)
report at least a two-fold increase of ETR103 in activated
lymphoblasts.
[0104] Oligonucleotide probes can also be ranked and/or selected by
base composition criteria that allow for more efficient synthesis
or preparation of the probes. For example and not by way of
limitation, in preferred embodiments of the present invention
oligonucleotide probes are selected for use on microarrays that are
prepared, e.g., by means of an ink jet printing device for
oligonucleotide synthesis (see, e.g., the methods and systems
described by Blanchard in International Patent Publication No. WO
98/41531, published Sep. 24, 1998; Blanchard, U.S. Pat. No.
5,028,189 issued Feb. 22, 2000; Blanchard et al., 1996, Biosensors
and Bioelectronics 11:687-690; and Blanchard, 1998, in Synthetic
DNA Arrays in Genetic Engineering, Vol. 20, J. K. Setlow, Ed.,
Plenum Press, New York at pages 111-123). In such embodiments, as
nozzles in the inkjet mechanism age, their firing accuracy
decreases. In a particular embodiment, nozzles in the inkjet
mechanism provide tetrazole activator to pairs of phosphoramidites
in combination. Thus, cytosine and adenine receive tetrazole
activator from one set of such nozzles while guanine and thymine
receive tetrazole activator from another set of such nozzles. As a
consequence, aging and/or misfiring nozzles can pull all available
cytosine and adenine bases to one side of the spot on the
microarray wherein the oligonucleotide probe is being synthesized.
Consequently, sequences that are substantially richer in either
adenine or cytosine can be synthesized with less purity and having
low-complexity sequences (e.g., homopolymers of adenine or
cytosine) at the fringe of the spot where they are synthesized on
the microarray. In such embodiments, therefore, it is preferable to
select probes for which the difference in the percentages of
adenine and cytosine bases ("A-C %") is very low or zero.
[0105] Exemplary, preferred base compositions values are as
follows; G+C %: preferably 0-75%, more preferably 0-55%, still more
preferably 8-45%; G %: 0-35%; C %: 0-35%; A %: 0-90%; T %: 0-90%;
A-C %: -15 to 60%; T-G %: -15 to 60%.
Sequence Complexity and Information Content:
[0106] As is readily appreciated by those skilled in the art,
oligonucleotide probes that hybridize most specifically to a
particular polynucleotide sequence (e.g., the sequence of a
particular gene) are, in general, probes that are complementary to
unique portions of the polynucleotide sequence that are not found
in other polynucleotide sequences (e.g., in other genes) in a given
sample. Such sequences are said to have a high information content
since they can identify unique polynucleotide sequences in a given
sample (for example, a unique gene in the genome of a cell or
organism). Conversely, probe sequences said to have low information
content are sequences whose complements can be found many times in
a sample (e.g., in the genome of a cell or organism) and which do
not, therefore, identify a unique polynucleotide such as a unique
gene.
[0107] Examples of sequences having a low information content
include, but are not limited to, repetitive elements, simple
repeats, and runs of contiguous repetition or "runs" of one base.
Contiguous runs of a single base are referred to in the art as
"polyX" runs or "polyX" repeats, wherein "X" denotes the nucleotide
base (e.g., adenine, thymine, guanine or cytosine) that is
repeated. Such polynucleotide repeats can be "scored" in a probe
sequence, e.g., by simply counting the number of nucleotide bases
in the single longest continuous run of any one base or,
alternatively, by totaling the cumulative length of bases involved
in polyX runs in the probe sequences. For example, when target
polynucleotide samples are prepared by a method comprising oligo-dT
priming of polyA+mRNA, a high proportion of polyT sequences may be
found at the 3' ends of the resulting polynucleotide molecules.
Accordingly, in such embodiments it is preferable to select probes
having few or no polyA repeats. Probes can therefore be evaluated
or scored for polyA repeats (or for continuous runs of any other
particular nucleotide base) by counting the number of contiguous
adenines (e.g., at the 5' end) and ranking or selecting the probes
so that probes having a lower polynucleotide repeat score are
preferably selected. polyX runs can be as short as two bases.
However, polyX runs that are more than three, four, five or ten
bases in length have particularly low information content and are
preferably avoided in oligonucleotide probes ranked and/or selected
according to the present invention.
[0108] "Simple repeats" refer to tandem repeats of short (e.g., 1-5
bases, more typically 1-3 bases) sequences. By contrast, repetitive
elements are longer (e.g., between 20 and 90,000 base pairs, more
typically about 1,000 base pairs), more complex sequences that are
overrepresented in a polynucleotide sample. For example, it is well
known in the art that the genomes of many higher organisms,
particularly eukaryotes (in particular, higher eukaryotes such as
mammals and including humans) contain complex sequences that occur
many times and are overrepresented in the genome. Typically, these
complex repeated elements are specific to the evolutionary lineage
of the cell or organism.
[0109] Both simple repeats and more complex repetitive elements can
be readily identified and "scored" by the skilled artisan. For
example, in a preferred embodiment, the program RepeatMasker
(Available Web Site:
http://ftp.genome.washington.edu/cgi-bin/RepeatMasker) can be used
to compare a polynucleotide sequence of interest (which is usually
entered by a user) to sequences of repetitive elements and/or
simple repeats in a database of such sequences. Because such
repetitive elements and simple repeats are generally specific to
the species of organism from which a polynucleotide sample is
derived, preferably the database is a database of repetitive
elements and/or simple repeats for an appropriate organism or class
of organism (e.g., for primates, rodents, mammals, vertebrates,
Arabidopsis, grasses or Drosophila, to name a few). Typically, such
a comparison is done using a "scoring matrix" that can be entered
or selected by a user or, alternatively, a default scoring matrix
used automatically by the program.
[0110] In a preferred embodiment, regions of the nucleotide
sequence of interest that align with repetitive element and/or with
simple repeat sequences within the database are "masked," e.g., by
replacing the aligned bases with "N" or "X" in the program output.
A skilled artisan can then select oligonucleotide probes with high
information content by selecting oligonucleotide sequences that are
complementary to portions of the target sequence that are not
masked.
Position:
[0111] The candidate polynucleotide probes evaluated according to
the methods and compositions of the present invention can be
complementary to any region of the target polynucleotide sequence
of interest (e.g., to any region of a gene sequence of interest).
For example, candidate polynucleotide probes having a nucleotide
sequence that is complementary to the nucleic acid sequence of a
particular target polynucleotide can be selected or provided by a
method that is referred to herein as "tiling." Specifically,
polynucleotide probes having a nucleotide sequence of length l are
selected by selecting probes having a nucleotide sequence
complementary to a sequence of l consecutive bases of the target
sequence. For example, a polynucleotide probe can be selected or
provided by selecting or providing a polynucleotide probe having a
nucleotide sequence complementary to l consecutive bases of the
target polynucleotide sequence beginning at the i'th base of the
target polynucleotide sequence. Thus, a first polynucleotide probe
can be selected or provided by selecting or providing a
polynucleotide probe whose polynucleotide sequence is complementary
to the nucleotide sequence corresponding to bases i through i+l of
the target polynucleotide sequence. A second polynucleotide probe
sequence can be selected or provided by selecting or providing a
polynucleotide probe whose nucleotide sequence is complementary to
the nucleotide sequence corresponding to bases (i+n) through
(i+n)+l of the target polynucleotide sequence, and so forth.
[0112] As noted above, l specifies the length of the probe's
polynucleotide sequence. Therefore, l is a positive integer,
preferably having a value between 4 and 200, and more preferably
having a value between 15 and 150. In embodiments wherein probes
having shorter oligonucleotide sequences are used, l is preferably
less than 40, more preferably between 15 and 30. Most preferably,
however, probes having longer oligonucleotide sequences are used.
In such embodiments, l is preferably between 40 and 80, more
preferably between 40 and 70, more preferably between 50 and
60.
[0113] n, the "tiling interval," is a positive integer that
preferably has a value between 1 and about 10. Particularly
preferred values of the tiling interval include n=1, 2, 3, 4 and 5.
i, which indicates the starting position within the target
polynucleotide sequence, is also a positive integer. In certain
preferred embodiments, the starting position is at or near the
5'-end of the target polynucleotide sequence. Thus, i has preferred
values less than 50 and more preferably less than 10. The first
base in the target polynucleotide sequence is a particularly
preferred starting position in such embodiments. Accordingly, a
particularly preferred value of the starting position is i=1. In
other preferred embodiments, only the 3'-end of the target
polynucleotide sequence is tiled. For example, in certain
embodiments, only the last 2,000, more preferably the last 1,000,
more preferably the last 500 and even more preferably the last 350
bases on the 3'-end of the target polynucleotide sequence are
tiled. In such embodiments, the value of the starting position i is
adjusted accordingly (e.g., i=L-2,000; i=L-1,000; i=L-500; or
i=L-350; wherein L is the length of the target polynucleotide
sequence).
[0114] In most preferred embodiments, the target polynucleotide
samples are prepared by amplifying "template" polynucleotide
molecules (e.g., mRNA molecules extracted from cells), as described
in Section 5.3.4 to produce a sample of cDNA or cRNA molecules. In
such preferred embodiments, amplification of the template
polynucleotide molecules is generally initiated at one of the two
distinct ends of the template polynucleotide molecules: the 5'-end
or the 3'-end. Because such amplification techniques are less than
100% efficient, a portion of the sequence of the template
polynucleotide molecule that is closer to the end where
amplification is initiated is preferentially amplified and is
therefore present in the target polynucleotide sample in greater
abundance. By contrast, a portion of the sequence of the template
polynucleotide molecule that is further from the end where
amplification is initiated are less preferably amplified and is
therefore present in the target polynucleotide sample in lower
abundance and may even be very rare or absent in the target
polynucleotide sample.
[0115] In preferred embodiments, therefore, candidate probes are
ranked and/or selected according to the distance of their
complementary region in the target polynucleotide sequence from the
preferentially labeled end of the target polynucleotide sequence.
Specifically, candidate probes are ranked and/or selected so that
those candidate probes corresponding to complementary regions of
the target polynucleotide that are near the preferentially labeled
end are chosen over candidate probes corresponding to complementary
regions of the target polynucleotide that are far from the
preferentially labeled end. In one embodiment, for example,
oligonucleotide probes are selected for which the first nucleotide
base of the corresponding region of the target polynucleotide is
within a chosen distance, referred to herein as the "end-distance,"
from the preferentially labeled end.
[0116] The exact end-distance will depend on the specific
amplification technique used to generate the target polynucleotide
sample. Preferably, the end-distance used is the distance from the
end of the target polynucleotide sequence that is amplified with at
least 50%, 60%, 70%, 80%, 90%, 95% or 99% efficiency. Appropriate
values for the end-distance can be readily determined by the
skilled artisan, e.g., using values in the literature for
particular amplification techniques used or, alternatively, through
routine gel electrophoresis experimentation to determine the length
of amplified fragments.
[0117] In addition, because both the sensitivity and specificity of
probes for a target polynucleotide sequence will typically vary in
a continuous manner as one "tiles" through the target
polynucleotide sequence as described above, oligonucleotide probes
can also be ranked and/or selected on the basis of their overlap
with other candidate polynucleotide probe sequences. That is to
say, in certain embodiments candidate polynucleotide probes can be
ranked and/or selected according to the amount of sequence they
share with other candidate polynucleotide probe sequences for the
same target polynucleotide.
[0118] For example, in one preferred embodiment candidate probes
are first ranked according to one or more of the other properties
or parameters described herein (e.g., sensitivity, specificity,
perfect match binding energy, base composition, position, etc.).
The top ranked probe can then be selected and compared to the
second ranked probe. Specifically, the overlap between the two
probes can be evaluated, e.g., by comparing the starting position,
i, of each probe within the target polynucleotide sequence. If the
overlap between the two probes is above a selected threshold (e.g.,
if the starting positions differ by more than 2 nucleotide bases,
more preferably by more than 5, 10, 20, 30, 40, 50 or 60 nucleotide
bases) than the second probe is also selected. However, if the
overlap between the two probes is equal to or above the selected
threshold, the second probe is rejected and the next probe (i.e.
the third probe) is selected and its overlap with the first probe
is evaluated. This process can be repeated until all of the ranked
candidate probes available have been either selected or rejected
or, alternatively, until a specified number of probes have been
selected. The selected probes can then be employed, e.g., for use
on a microarray or, more preferably, can be further screened
according to other conditions and criteria discussed above.
5.1.3. ITERATIVE RANKING OF CANDIDATE PROBES
[0119] Candidate probes of the present invention can be ranked
according to a variety of ranking systems. Preferably, the systems
are based on at least two, and more preferably a plurality of the
properties and parameters described in Section 5.1.2, above. For
example, in preferred embodiments, candidate probes are ranked
according to both the sensitivity and the specificity with which
the probe hybridizes to a target polynucleotide sequence (e.g.
using a target binding energy score and a non-target binding or
cross hybridization energy score). However in more preferred
embodiments, the candidate nucleotide probes can, in fact, be
ranked and/or selected according to any combination of properties
and parameters described hereinabove, including but not limited to:
(a) probe size or length; (b) binding energies, including both the
perfect match duplex (i.e., of a probe and its target,
complementary nucleotide sequence) and cross-hybridization binding
energies; (c) base composition, including, for example, the
relative amount or percentage of one or more particular nucleotide
bases (e.g., adenine, guanine, thymine or cytosine) in a probe
sequence, as well as the relative amount or percentage of any
combination of such nucleotide bases; (d) the position of a probe's
complementary sequence in the sequence of its "target"
polynucleotide or gene sequence; and (e) probe sequence complexity,
including the presence or lack of common repetitive elements such
as polynucleotide repeats (i.e., simple, contiguous repeats of one
or more nucleotide bases) as well as more complicated repetitive
elements that are well known in the art. Still other exemplary
parameters which can be used in the methods and compositions of the
invention for ranking and/or selecting oligonucleotide probes
include: (f) self dimer binding energy (i.e., the tendency for a
particular probe to hybridize to its own sequence); (g) the
structure content of the complementary, target polynucleotide
sequence for a particular probe (e.g., the presence or absence of
certain structural features or motifs); and (h) the information
content of a probe's nucleotide sequence. Other properties and
parameters known in the art to influence or be predictive of
hybridization and cross-hybridization can also be used to rank
and/or select candidate nucleotide probes according to the methods
of the present invention.
[0120] As an example, and not by way of limitation, threshold
values (or ranges of acceptable values) can be selected for one,
two, three, four or more of the properties described, e.g., in
Section 5.1.2, above. Candidate probes can then be selected that
have values of those properties that are above (or below) the
thresholds or that are within the selected ranges. The selected
probes are then ranked according to some other property such as
their perfect-match binding energy scores or, alternatively, their
cross-hybridization binding energy scores (e.g. the "minmax" score
described in Section 5.1.2 above).
[0121] Alternatively, candidate probes can be ranked according to
each of two, three, four or more selected properties such as the
properties described in Section 5.1.2, above. A combined rank can
then be determined for each probe that is based, e.g., on the sum
of the individual rankings. Such a sum can be, for example, an
unweighted arithmetic sum or, alternatively, a weighted arithmetic
sum using appropriate weighting factors.
[0122] In yet another alternative embodiment, candidate probes can
also be ranked by, first, selecting candidate probes that have
values of one, two, three, four or more properties (e.g., selected
from the properties described in Section 5.1.2) that are each above
(or below) a selected threshold or which, alternatively, are within
a selected range of values. The selected probes can then be ranked
according to each of two, three, four or more selected properties
(e.g., from the properties described in Section 5.1.2) and a
combined rank, based, e.g., on the sum of the individual rankings,
can be determined for each probe.
[0123] In more specific, and non-limiting, exemplary embodiments,
given the sensitivity and specificity of each candidate probe, the
probes may then be ranked according to a variety of ranking systems
which will be readily apparent to those skilled in the relevant
art. For example, in one preferred embodiment, a threshold
sensitivity or perfect-match binding energy is selected, and those
probes whose sensitivity or perfect-match binding energy lies above
the threshold are ranked according to their specificity, i.e., so
that those probes above the threshold having the highest
specificity have the highest rank. Alternatively, an interval
(i.e., a range) of sensitivity or perfect-match binding energy
values can be chosen and probes within that interval can be ranked
according to their sensitivity. Conversely, a threshold specificity
value (or range of specificity values) may be used and those probes
whose specificity lies above that value (or within that range of
values) may be ranked according to their sensitivity or according
to their binding energies.
[0124] In an alternative preferred embodiment, the probes may be
ranked twice: once according to their sensitivity (or according to
their perfect-match binding energy) and once according to their
specificity. A combined rank may then be determined for each probe
which is based upon the sum of the sensitivity (or perfect-match
binding energy) rank and the specificity rank. In one aspect of
this embodiment, the sum of the sensitivity (or perfect-match
binding energy) rank and the specificity rank may be a weighted
sum, using appropriate weighting constants. One skilled in the
relevant art will readily appreciate how to select appropriate
values for such weighting constants depending upon the particular
circumstances (e.g., the particular polynucleotide molecules to be
analyzed and/or their relative abundances).
[0125] As an exemplary and non-limiting embodiment, FIGS. 7A-B each
plot the predicted binding energy score (i.e., the predicted
binding energy .DELTA.G) vs. the predicted cross hybridization
score (i.e., the predicted value of X.sub.p/T.sub.p) for the 10%
highest binding energy probes for two S. cerevisiae genes: YER019W
(FIG. 7A), and YER019C (FIG. 7B). In particular, the binding energy
score was predicted using the nearest-neighbor model (i.e., from
Equation 1, above) and the values of T.sub.p and X.sub.p were
evaluated from Equations 8 and 9, respectively. SAGE abundance
estimates (see, Velculescu et al., 1995, Science 270:484-487; and
Velculescu et al., 1997, Cell 88:243-251) were used to evaluate the
abundance term M(c) in Equations 8 and 9. Both the binding energies
and the specificity values were normalized to have zero mean and
unit variance.
[0126] The sensitivity and specificity values of probes for the
gene YER019W (FIG. 7A) are typical of those obtained for genes that
are fairly unique within a sample, e.g., such as genes that have no
close homologs or analogs in the genome of an organism. In
particular, probes which have both high binding energy and high
specificity can be readily identified, e.g., by visual inspection
of FIG. 7A. In such embodiments, ranking systems that select probes
having a minimum required specificity (or having a specificity
within some range of specificity values) and rank the selected
probes according to their sensitivity will be preferred. By
contrast, the distribution of sensitivity and specificity values of
probes for the gene YAR010C (FIG. 7B) are typical of genes that are
members of homology families and to which there are typically
similar cross-hybridizing sequences in a sample. In such an
embodiment, it is readily difficult to identify probes having both
high specificity and high sensitivity by visual inspection of FIG.
7B alone. In embodiments such as this, ranking systems are
preferred that select probes having a minimum required sensitivity
(or having a sensitivity within some range of sensitivity values)
and rank the selected probes according to their specificity.
[0127] In another exemplary and non-limiting embodiment, the
hybridization intensity (i.e., brightness) vs. specificity of the
S. cerevisiae genes YER019W and HXT3, respectively, are plotted in
FIGS. 8A-B using the experimental hybridization data plotted in
FIGS. 5 and 6 and discussed above. Those probes which were
predicted to rank highest based on the above-discussed ranking
functions are indicated in FIGS. 5B and 6B by an (X) symbol.
Specifically, the probes were ranked based on a combined ranking
function in which both probe specificity and sensitivity were
weighted equally. As can be seen in FIGS. 8A-B, the predicted top
ranking probes do indeed tend to have higher sensitivity and
specificity.
[0128] In addition to the analytical ranking systems described
above, candidate probes may also be ranked according to empirical,
iterative methods. Most preferably, the candidate probes of the
invention are ranked according to both analytical and empirical
ranking systems and/or methods. In particularly preferred
embodiments, candidate oligonucleotide probes are first ranked
according to the above-described analytical methods, and such
ranking is then empirically refined, at least for the highest
ranked probes.
[0129] For example, candidate oligonucleotide probes, such as high
ranking candidate probes for one or more target polynucleotides,
may be empirically ranked by synthesizing one or more microarrays
comprising the candidate probes (103) and hybridizing a reference
polynucleotide sample thereto (104). Preferably, such hybridization
occurs under conditions such as those described in Section 5.1.2,
above, so that hybridization intensity (i.e., hybridization signal
intensity) correlates with probe specificity. Thus, by empirically
selecting for probes with high hybridization intensity (105), the
candidate probes are selected for both sensitivity and
specificity.
[0130] An exemplary and more detailed embodiment of the ranking
methods of the invention is shown in FIG. 19. In this particular
embodiment, oligonucleotide probes are ranked and (optionally)
selected for detecting a particular polynucleotide sequence
selected by a user. Usually, the polynucleotide sequence will be
the sequence of a gene that is expressed (or suspected of being
expressed) by a cell or organism.
[0131] Optionally, the complexity or information content of the
polynucleotide sequence is first analyzed using a program such a
RepeatMasker, described above, to identify portions of the sequence
that have low information content such as, but not limited to,
portions of the sequence corresponding to repetitive elements or
simple repeats. Next, a maximum distance (e.g., a maximum number of
nucleotide bases) from the 3'-end of the polynucleotide sequence is
selected, and only the portion of the polynucleotide sequence
within this selected distance from the 3'-end is further analyzed.
Oligonucleotide sequences are then generated, e.g., according to
the tiling methods described hereinabove, having a particular
sequence length (or, alternatively, a particular range of sequence
lengths) that is usually selected by a user.
[0132] These oligonucleotide sequences are then evaluated as
candidate probe sequences. First, those oligonucleotide sequences
corresponding to regions of low information content in the target
polynucleotide sequence are removed from consideration. In
particular, oligonucleotide sequences that contain, e.g., all or
part of a repetitive element or a simple repeat identified by a
program such as RepeatMasker are removed. Likewise, oligonucleotide
sequences having one or more polyX repeats that are greater than a
particular length (e.g. greater than 2, greater than 3, greater
than 4, greater than 5, greater than 6, greater than 7, greater
than 8, greater than 9, greater than 10, greater than 15 or greater
than 20) are removed from consideration. In addition, those
oligonucleotide sequences that correspond to unknown or variant
sequences of the target polynucleotide sequence (e.g., where one or
more allelic variants of the target polynucleotide sequence are
known to exist) are also preferably removed from consideration.
[0133] The base composition (e.g., G+C %, G %, C %, A %, T %, A-C
%, T-G %, etc.) of the candidate sequences is also preferably
evaluated. Exemplary, preferred base composition values are as
follows; G+C %: preferably 0-75%, more preferably 0-55%, still more
preferably 8-45%; G %: 0-35%; C %: 0-35%; A %: 0-90%; T %: 0-90%;
A-C %: -15 to 60%; T-G %: -15 to 60%.
[0134] The sequences for candidate oligonucleotide probes having
been obtained, the perfect-match duplex binding (i.e.,
hybridization) energy .DELTA.G.sub.p is than calculated for each
candidate oligonucleotide probe p using formulas well known in the
art, such as the formulas of the nearest neighbor model and, in
particular, Equation 1, above. Optionally, candidate
oligonucleotide probes having a calculated value for .DELTA.G.sub.p
that is below a certain threshold or that lies outside a certain
range of values are removed from consideration and are not further
evaluated. Typically the threshold or range of values will be a
range or threshold selected by a user, and one skilled in the art
can readily select appropriate values without undue
experimentation. Exemplary threshold binding energy values include,
e.g., 100, 60 or 23 kcal/mol.
[0135] The remaining candidate oligonucleotide probes are then
ranked, first according to length (i.e., number of nucleotide
bases), and second according to the distance of the probe's
nucleotide sequence from the 3'-end of the target polynucleotide
sequence. The ranked candidate probes are then "de-overlapped" to
select those probes whose sequences overlap the target
polynucleotide sequence by no more than a certain number of
nucleotide bases selected by the user (e.g., by no more than 2, 5,
10, 30 or 60 nucleotide bases). Such de-overlapping can be
performed, e.g., according to the methods described hereinabove.
Specifically, the top ranked candidate probe can first be selected.
The next ranked candidate probe whose sequence overlaps the
sequence of the first selected probe by no more than the specified
number of bases is also selected, and so forth. Those candidate
probes that are not selected are then removed from further
consideration.
[0136] Preferably, the probe chosen for a given target
polynucleotide will, in the worst case scenario, be the probe with
the least or least objectionable amount of hybridization with other
polynucleotide sequences in a sample. Thus, in preferred
embodiments wherein the probes are used to detect expression of
particular genetic transcripts from the genome of a cell or
organism, the chosen probe will be the probe that hybridizes least
favorably (i.e., with the least negative binding energy) to the
other sequences in the genome of that cell or organism. Such a
probe can be identified, for example, by means of a homology search
method such as BLAST (Altschul et al., 1990, J. Mol. Biol.
215:403-410; Altschul, 1997, Nucleic Acids Res. 25:3389-3402; and
Zhang and Madden, 1997, Genome Res. 7:649-656). In particular, a
BLAST search can be performed against each candidate probe sequence
to identify polynucleotide sequences other than the target
polynucleotide sequence that are identical or homologous to the
probe sequence. For example, sequences that are at least 50%, 60%,
70%, 80%, 90%, 95%, 99% or 100% identical to a candidate probe
sequence are identified using a search algorithm such as BLAST.
Preferably, the database of sequences used in such an identity or
homology search is a database of or containing all or substantially
all of the sequence that are present or believed to be present in a
polynucleotide sample that the candidate probe or probes are
intended to assay. Thus, for example, in embodiments wherein the
target polynucleotide is a gene expressed by a particular cell or
organism, the database is preferably a database of or containing
all or substantially all of the gene sequences expressed by that
cell or organism, or a database of all or substantially all of the
gene sequences in the genome of that cell or organism. In preferred
embodiments, for example, the database may contain at least 50%, at
least 60%, at least 70%, at least 80%, at least 90%, at least 95%,
at least 99% or 100% of the sequences expressed by the cell or
organism. The database can also be a database containing at least
50%, at least 60%, at least 70%, at least 80%, at least 90%, at
least 95%, at least 99% or 100% of the gene sequences in the genome
of the cell or organism. Publicly available databases of expressed
sequences can also be used in such an analysis including, for
example, the GenBank or dbEST databases. In addition, because many
of the records in such databases are, in fact, duplicate records of
the same gene, a cluster filed such as the UniGene cluster file
(Schuler, 1997, J. Mol. Med 75:694-698; Schuler et al., 1996,
Science 274:540-546; Boguski & Schuler, 1995, Nature Genetics
10:369-371) can also be used to identify matches to polynucleotide
sequence that are, in fact, the target polynucleotide sequence.
[0137] Candidate oligonucleotide probes that are found to have 100%
sequence identity to a sequence that is not part of the target
polynucleotide sequence (i.e., candidate probes that have other
perfect-match sequences in the database of expressed sequences) are
preferably rejected and eliminated from further analysis or
consideration. However, such probes can be used in certain
embodiments, e.g., as probes for a family or families of genes
whose members share a common sequence or common sequences. For
those sequencesj in the database that only partially align with the
candidate probe sequences (e.g., sequences that are at least 50%,
at least 60%, at least 70%, at least 80%, at least 90%, at least
95% or at least 99% identical to a candidate probe sequence), the
cross-hybridization binding energy .DELTA.G.sub.i of these
sequences to the candidate probe is also calculated according to
the methods described supra for the perfect match binding energy
(e.g., using the nearest neighbor model and Equation 1, above). In
a particularly preferred embodiment, the most negative
cross-hybridization binding energy, which represents a "worst-case"
cross-hybridization scenario, is identified and this value is used
as the cross-hybridization score for the candidate oligonucleotide
probe. In a less preferable embodiment, however, the level of
"target" hybridization T.sub.p of the candidate oligonucleotide
probe p to the target polynucleotide sequence and the level of
cross hybridization X.sub.p of the candidate oligonucleotide probe
to non-target sequences (e.g., the sequences identified by BLAST or
another sequence homology search algorithm) can be calculated,
e.g., according to Equations 8 and 9, above, and used as the
cross-hybridization score for the candidate oligonucleotide probe.
The remaining candidate oligonucleotide probes can then be
re-ranked according to their cross-hybridization scores. Those
candidate probes having the the most positive cross-hybridization
scores are preferably selected for use, e.g. in the microarrays of
the present invention. Preferably, the selected probes are
candidate probes that have, not only low cross-hybridization
scores, but also have a high perfect-match binding energy.
[0138] In preferred embodiments, new microarrays are prepared using
the highest ranking probes identified by the above-described
methods, and steps 103-105 of FIG. 1 are iteratively repeated to
identify those probes with the highest sensitivity and specificity.
For example, steps 103-105 of FIG. 1 may be repeated until the
number of probes per target polynucleotide has been reduced to some
upper limit, e.g., so that the candidate probes for all target
polynucleotides may be incorporated on a single microarray (i.e.,
on a single "chip"). Alternatively, probes may be iteratively
ranked according to the above-described methods until some other
criteria is satisfied, such as when the selected probes satisfy a
minimum (e.g., user determined) required sensitivity and
specificity.
[0139] In one preferred embodiment, the number of candidate probes
for each target polynucleotide is no more than 20, more preferably
no more than 10, and more preferably no more than 5, no more than
4, no more than 3, or no more than 2. In particularly preferred
embodiments, one candidate probe is identified and/or used for each
target polynucleotide.
5.1.4. SCREENING CHIPS
[0140] The present invention also provides "screening chips" which
comprise probes for a large number of different polynucleotides. As
used herein, a "chip" comprises a single microarray of
polynucleotide probes bound to a solid support. The solid support
may be a porous or non-porous support. Microarrays are well known
in the art and are described in detail in Section 5.3 below.
[0141] In particular, the screening chips of the invention are able
to detect, by hybridization, expressed polynucleotide sequences
(e.g., mRNA of expressed genes or cDNA derived therefrom)
representing the entire genome of a cell or organism. The screening
chips of the invention preferably comprise probes that hybridize
specifically and distinguishably to at least 50% of the genes in
the genome of a cell or organism. More preferably, the screening
chips comprise probes that hybridize specifically and
distinguishably to at least 75%, at least 80%, at least 85%, at
least 90%, at least 95% or at least 99% of the genes in the genome
of a cell or organism. In a particularly preferred embodiment, the
screening chips comprise probes that hybridize specifically and
distinguishably to all (i.e., 100%) of the genes in the genome of a
cell or organism In other embodiments, however, the screening chips
have probes for those particular genes expressed by a particular
cell or cell type of interest. In such embodiments, a screening
chip will therefore preferably have probes that hybridize
specifically and distinguishably to all of the genes expressed by
the cell or cell type of interest, which will often be
substantially less than 50% of the genes in the entire genome of
the cell or organism (e.g., 20%). The organism may be of any
species, including procaryotic organisms, such as E. coli and other
bacteria, and eukaryotic organisms including, but not limited to,
Saccharomyces cerevisiae. The organism may also be a higher,
multi-cellular organism such as a plant or animal, including a
mammalian animal such as a mouse or a human.
[0142] In particularly preferred embodiments, the screening chips
of the invention include probes for all of the expressed
polynucleotide sequences (i.e., for the entire genome) of a cell or
organism. In preferred embodiments, therefore, the screening chips
contain probes that can hybridize specifically and distinguishably,
and can therefore detect, at least about 2000 or at least about
4000 polynucleotide sequences. More preferably, the screening chips
contain probes to detect at least about 10,000, at least 15,000, or
at least about 20,000 polynucleotide sequences. In particularly
preferred embodiments, the screening chips contain probes to detect
greater than 80,000, greater than 100,000 or greater than 150,000
polynucleotide sequences.
[0143] The screening chips of the invention maximize the number of
polynucleotides that may be detected by minimizing the number of
probes needed to detect each polynucleotide sequence. In
particular, by selecting probe sequences according to the methods
and/or having the lengths disclosed hereinabove, the number of
probe sequences required to report a particular polynucleotide
sequence (e.g., the sequence of a particular gene or gene
transcript) may be reduced to as few as one probe sequence. The
probe sequences used in the screening chips hybridize specifically
and distinguishably to a particular target polynucleotide sequence
such as the sequence of a particular gene or gene transcript. Thus,
the amount of cross hybridization to other sequences is minimized.
In fact, in preferred embodiments the amount of cross hybridization
is zero, or is at least negligible. Thus, the amount of
hybridization to a particular probe is a reliable indicator of the
relative amount of a particular target polynucleotide sequence
present in a sample. More specifically, although absolute
hybridization intensity values for a target polynucleotide sequence
within a sample may vary among different probe sequences, changes
(e.g., ratios) of the hybridization levels between perturbed and
unperturbed cells (for example, between cells exposed to a drug and
cells that are not drug-exposed) are consistent among individual
probes. Thus, changes in gene expression can be accurately and
reliably measured using a single probe.
[0144] This principle is illustrated by example in FIG. 11 which
shows the correlation between data obtained using a screening chip
of the present invention and data from a conventional microarray.
Specifically, otherwise identical cultures of S. cerevisiae were
left untreated or were treated with 10 mM
3-amino-1,2,4-triazole(3-AT). The poly A+fraction of total cellular
RNA from each culture was isolated and amplified by in vitro
transcription ("IVT") as described by van Gelder et al. (U.S. Pat.
No. 5,716,785). IVT products from the drug-treated cell culture
were labeled with Cy5, whereas those from the untreated cell
culture were labeled with Cy3 according to standard protocols. The
labeled samples were hybridized to a screening chip, prepared
according to the above-described methods, comprising a single
oligonucleotide probe sequence specific for each gene in the yeast
genome. Labeled samples were also hybridized, in parallel, to a
conventional microarray (a Y36100 Set GeneChip.RTM. Yeast
Expression Analysis Product from Affymetrix, Santa Clara, Calif.)
which comprised 40 oligonucleotide probe sequences for each gene of
the yeast genome (20 match sequence probes and 20 mismatch sequence
probes). Each of the hybridized chips was scanned using a laser
confocal scanner or an Affymetric GeneChip.RTM. instrument system,
respectively. A distribution plot of the hybridization ratio
between treated and untreated cells is shown in FIG. 11.
Specifically, the plot compares the expression ratios measured with
the screening chip (horizontal axis) and the conventional
microarray (vertical axis). The high correlation coefficient
r=0.85) obtained demonstrates that the observed expression ratios
are very similar for both data sets.
[0145] FIG. 12 shows exemplary data demonstrating the measured
expression ratios are consistent among different probes even though
absolute intensities of hybridization measured for each probe may
vary. In particular, cultures of a wild-type strain of S.
cerevisiae and a strain having a homozygous diploid deletion
mutation in the pep12 gene were harvested and RNA was isolated and
amplified and labeled with Cy3 (wild type) and Cy5 (pep12 deletion)
according to the above described methods, and hybridized to a
screening chip comprising nine different oligonucleotide probe
sequences for each gene of the S. cerevisiae genome. The average
and standard deviation of intensities and expression ratios for the
probe sequences for each gene were determined and a fractional
error (standard deviation divided by the mean) calculated. FIG. 12
shows a histogram of the distribution of fractional errors for the
absolute hybridization intensities (dashed line) and expression
ratios (solid line) from this data. Fractional errors of absolute
hybridization intensities were greater than fractional errors of
expression ratios, as can be seen in FIG. 12, suggesting that a
single oligonucleotide probe sequence can be used to accurately
report changes in gene expression
[0146] In preferred embodiments, the screening chips contain no
more than 10 probes for each target polynucleotide sequence. More
preferably, the screening chips contain no more than 5, no more
than 4, no more than 3, or no more than 2 probes for each target
polynucleotide sequence. Most preferably, the screening chips
contain only one probe for each target polynucleotide sequence.
Accordingly, the probes used in the screening chips must be
optimized for sensitivity and specificity, and are thus most
preferably selected according to the methods described above in
Sections 5.1.1-5.1.3.
[0147] The screening chips of the invention are particularly useful
for identifying "signature genes" of a cell or organism, i.e.;
genes whose expression changes in response to particular changes or
perturbations to that cell or organism. Screening chips can
therefore be used, e.g., to identify the genes of a cell or
organism whose expression is up-regulated or down-regulated, e.g.,
as a result of exposure to one or more drugs or to a particular
class or family of drugs, as a result of a mutation and/or a change
in the expression of one or more other genes (e.g., using a
controllable promoter such as a titratable promoter), or as a
result of changes in the cell or organism's environment including
changes in temperature, exposure to moderate doses of radiation and
changes in the nutritional environment, such as the presence or
absence of certain sugars or amino acid residues, to name a
few.
[0148] The identification of such signature genes is therefore
useful in a variety of applications and methods for characterizing
cells and organisms, including testing biological network models
(see, in particular, U.S. Pat. No. 6,132,696), identifying pathways
of drug action (U.S. Pat. No. 5,965,352), drug screening methods
(see, e.g., International Patent Publication WO 98/38329, published
Sep. 3, 1998), determining protein activity levels (U.S. patent
application Ser. No. 09/303,082 filed on Apr. 30, 1999), monitoring
disease states and therapies (U.S. Pat. No. 6,218,122) including
determining the therapeutic index of a drug (U.S. Pat. No.
6,222,093), and identifying drug targets (U.S. Pat. No. 6,146,830)
to name a few.
[0149] The screening chips of the invention can also be used, e.g.,
to identify signature genes that correspond to one or more
co-varying sets of genes or gene transcripts. That is to say, the
screening chips can be used to identify sets of genes or gene
transcripts which change together, e.g., by increasing or
decreasing their abundances and/or activities, under some set of
conditions. Such co-varying genes and gene transcripts include
genes and/or gene transcripts that are co-regulated including, for
example, genes or gene transcripts that share one or more
regulatory elements such as common regulatory sequence patterns.
For a detailed description of co-regulated and co-varying gene
sets, including methods for identifying co-regulated and
co-regulated genesets, see, e.g., U.S. Pat. No. 6,203,987; U.S.
Pat. No. 6,950,752; and U.S. Pat. No. 6,801,859.
[0150] Methods for measuring hybridization of polynucleotides to a
microarray, which are particularly suitable for identifying
signature genes in the methods of the present invention, are also
provided below in Section 5.3.6.
5.1.5. SIGNATURE GENES AND CHIPS
[0151] As noted above, the methods and compositions of the present
invention are particularly useful, e.g., for identifying "signature
genes" of a cell or organism; i.e., genes whose expression changes
in response to particular changes or perturbations to that cell or
organism. In particular, although the screening chips may confer up
to .about.10.sup.5 genetic transcripts, in most instances the
expression levels of a large part or even a majority of these
constituents will not change significantly in response to a
particular change or perturbation to the cell or organism, or the
change may be small and dominated by experimental error. This point
is illustrated, by example, in FIG. 9 which provides a
representation of the changes in abundances of 4000 genes as a
result of each of 350 different changes or perturbations to cells
of S. cerevisiae. Grey indicates no measurable change in the
abundance of a gene transcript, whereas black and white indicate an
increase or decrease, respectfully, in the abundance of a
particular gene transcript, indicated by the "gene index," in
response to a particular change or perturbation, which is indicated
by the "experiment index." It is generally unhelpful and cumbersome
to use these transcripts for most applications, including the
applications recited in Section 5.1.4 above. Preferably, therefore,
only those genetic transcripts whose abundances do change
significantly in response to changes or perturbations to the cell
or organism are examined. Such genetic transcripts are referred to
herein as "signature genes."
[0152] Signature genes are identified, in the methods of the
present invention, as those transcripts whose expression changes
beyond a selected threshold. For example, in most embodiments
changes in the hybridization between untreated and treated cells
are quantified in terms of log expression ratios. Thus, in one
embodiment a gene may be identified as part of a signature if its
log expression ratio is greater than or equal to a factor of two.
Alternatively, if error estimates can be derived for the expression
ratios, a confidence or probability value may be assigned to each
expression ratio representing the probability that it arose by
chance or in the absence of any actual change in expression (see,
e.g., U.S. Pat. No. 6,351,712) a threshold confidence or
probability values, e.g., 95% probability, can be used to define
the signature genes.
[0153] In those embodiments of the invention wherein signature
genes are identified using more than one probe per gene sequence,
the hybridization intensities for each probe are preferably
combined to generate an estimate of the target gene expression
levels and/or changes therein. For example, background additive
intensity errors can be estimated and subtracted from the
hybridization signal, e.g., using the average of negative control
probes as an estimate of the background signal, and the
hybridization intensities for the probes for a particular target
polynucleotide sequence can be averaged or otherwise combined
(e.g., additively) to provide a representation of the target
polynucleotide's expression level. In one embodiment, outlier
rejection is performed before such averaging or combining to remove
those signals that vary by more than a certain threshold a (e.g.,
by more than two to three standards of deviation from the primary
or average value). In another embodiment, which is discussed in
detail below, each polynucleotide probe occurs in pairs wherein the
second member of each pair is an intentional sequence variant
(i.e., a mismatch) of the first member. Accordingly, the
hybridization intensity of the second member provides an estimate
of the level of cross-hybridization to the first member.
Subtracting the hybridization intensity of the second member from
the intensity of the first member thus provides a correction for
the specific hybridization to the first member, at least to the
extent that the mismatch variant truely represents the
cross-hybridization level.
[0154] Once signature genes have been identified for particular
changes or perturbations to a cell or organism, probes for
detecting polynucleotide molecules corresponding to the signature
genes may be selected (108), e.g., using the ranking and selection
methods described in Sections 5.1.2-5.1.3, above, and "signature
chips" may be constructed (109) with the selected probes according
to standard methods for fabricating microarrays, including the
methods described in Section 5.3 below. Such signature chips are
therefore also considered part of the present invention.
[0155] The signature chips of the invention therefore comprise
arrays of polynucleotide probes which are selected according to
optimal sensitivity and selectivity for a particular set of
signature genes. Because the signature chips of the invention
contain probes for detecting fewer target polynucleotide sequences
(e.g., for fewer genes or gene transcripts), such chips can
accommodate a larger number of probes per target polynucleotide
sequence. For example, in preferred embodiments the signature chips
of the invention comprise at least five probes specific to each
target polynucleotide sequence. More preferably, the signature
chips contain at least 10 probes specific to each target
polynucleotide sequence or, in certain preferred embodiments, at
least 20 probes specific to each target polynucleotide sequence
(i.e., for each gene or gene transcript). In other preferred
embodiments, the signature chips of the invention contain at least
50, at least 100, at least 150 or at least 200 probes specific to
each target polynucleotide sequences. This redundancy among the
probes of a signature chip can be used to estimate and subtract
contributions to the hybridization intensity signal which are due
to cross-hybridization and thereby detected hybridization to a
particular target polynucleotide sequence (e.g., to a particular
gene) more accurately. Thus, the signature chips of the invention
are able to detect the actual level of particular polynucleotide
sequences in a sample (e.g., the actual level of expression of
particular genes or gene transcripts) more accurately than are
screening chips.
[0156] In one exemplary but non-limiting embodiment the signature
chips comprise both match and mismatch probes for a signature gene.
Methods for detecting polynucleotides using systems of
matched/mismatched probe pairs are known in the art, and are
described, e.g., by Lockhart et al., 1996, Nature Biotechnology
14:1675-1680. Specifically, one probe in each pair of probes is a
matched sequence probe that is matched to (i.e., complementary to)
and therefore specific for a particular target polynucleotide
sequence. The other probe in each pair of probes is an intentional
mismatch sequence probe which is not matched or complementary to
the target polynucleotide sequence of the matched sequence probe,
but which does have the same or about the same melting temperature
as the melting temperature of the matched sequence probe (i.e.,
within 5.degree. C. or, more preferably, within 2.degree. C.). For
example, in preferred embodiments, a mismatch sequence probe will
have between one and 3 single base mismatches to its target
polynucleotide sequence. Specifically, in embodiments wherein
shorter oligonucleotide probes are used (e.g., less than or equal
to about 30 bases in length) single base mismatches are preferred,
whereas double or triple base mismatches are preferred for longer
oligonucleotide probes (e.g., 50 to 60 base pairs in length or
longer).
[0157] Averaged over all possible cross-hybridizing sequences, the
mismatch and match probes will each have the same intensity from
cross-hybridization. Thus, the difference in signal intensity
between the match and intention mismatch probe of a particular pair
of probes is, in the mean, the hybridization intensity from
specific hybridization of the matched sequence probe to the target
polynucleotide sequence.
[0158] One skilled in the art will appreciate, however, that the
actual distribution of cross-hybridizing sequences in a real sample
may, in fact, have more over-all homology for one probe in a
match/mismatch pair of polynucleotide probes than for the other
probe. As a result, there will still be some random amount of
signal due to cross hybridization. Accordingly, in preferred
embodiments a plurality of match/mismatch probe pairs are used to
detect a single signature gene. As signals from the plurality of
match/mismatch probe pairs are combined (e.g., averaged), the
contribution of cross-hybridization to the combined signal will
decrease as the number of probe pairs increases. In particular, the
contribution of cross-hybridization to the combined signal will
tend to zero in the limit of a large number of probe pairs. For
example, in one preferred embodiments, 20 or more probe pairs are
used to detect a single signature gene.
[0159] In another, particularly preferred embodiment, set of
match/mismatch probe sequences are used. Specifically, each set
comprises a match sequence probe for a particular target
polynucleotide sequence and a plurality of mismatch sequence
probes. For example, in one exemplary embodiment, between 10 and
200 match sequence probes may be used on a signature chip that are
specific to a particular target polynucleotide sequence such as the
sequence of a particular gene. In such an embodiment, the signature
chip may also contain as many as 4 to 20, or more mismatch sequence
probes for each match sequence probe on the signature chip.
[0160] In another exemplary, but also non-limiting embodiment, a
large number of matched (i.e., complementary) probe sequences
(e.g., preferably between 5-10 or more probe sequences) may be used
to detect each gene rather than matched/mismatched probe pairs. In
such an embodiment, the amount of a signature gene present in a
sample is preferably determined by selecting the probes with the
highest hybridization intensity and combining (e.g., averaging)
their signals. For example, signature probes may be selected, e.g.,
by outlier rejection, whose hybridization intensities vary from the
mean hybridization intensity by no more than some threshold (e.g.,
some multiple of fraction of the standard deviation). In
particular, by hybridizing polynucleotides to the signature chips
of the present invention under the highly stringent conditions
discussed in Section 5.1.1, above, hybridization specificity will,
in general, correlate with specificity. Thus, those probes having
the highest hybridization intensity will generally be the probes
which hybridize most specifically to a target polynucleotide
sequence. Thus, the contribution of cross-hybridization to the
signal will be minimized. However, the contribution of
cross-hybridization to the combined (e.g., averaged) signal will
not tend to zero in the limit of a large number of probes. However,
this second exemplary embodiment is preferred in instances wherein
it is more preferable to have small variance than it is to have
small bias in hybridization measurements, whereas the first
exemplary embodiment is preferred in instances wherein it is more
preferable to have small bias than to have small variance in
hybridization measurements. For example, in those embodiments
wherein there is a large number (e.g., about ten or more pairs) of
probes per signal gene, the unbiased match/mismatch embodiment is
generally preferred, whereas in those embodiments wherein there is
a relatively small number of probes per signal gene (e.g., less
than about fifteen to twenty), the second exemplary embodiment is
preferred.
5.2. ANALYTICAL SYSTEMS
[0161] The analytic methods described in Section 5.1 above can
preferably be implemented by use of computer systems such as those
described herein. FIG. 10 illustrates an exemplary computer system
suitable for implementation of the analytic methods of this
invention. Computer system 1001 is illustrated as comprising
internal components and being linked to external components. The
internal components of this computer system include processor
element 1002 interconnected with main memory 1003.
[0162] It is noted that although the present description and
figures refer to an exemplary computer system having a memory unit
and a processor unit, the computer systems of the present invention
are not limited to those consisting of a single memory unit or a
single processor unit. Indeed, computer systems comprising a
plurality of processor units and/or a plurality of memory units
(e.g., having a plurality of SIMMS or DRAMS) are well known in the
art. Indeed, such systems are generally recognized in the art as
having improved performance capabilities over computer systems that
have only a single processor unit or a single memory unit. For
example, in one preferred embodiment, computer system 1001 is an
Alta cluster of nine computers; a head "node" and eight sibling
"nodes," each having an i686 central processing unit ("CPU"). In
addition, the Alta cluster comprises 128 Mb of random access memory
("RAM") on the head node and 256 Mb of RAM on each of the eight
sibling nodes. Nevertheless and as the skilled artisan readily
appreciate, as such computer systems relate to the present
invention, a computer system that has a plurality of memory units
and/or a plurality or processor units is, in fact, substantially
equivalent to the exemplary computer system depicted in FIG. 10 and
having only a single processor and a single memory unit.
[0163] The external components include mass storage 1004. This mass
storage can be one or more hard disks which are typically packaged
together with the processor and memory. Such hard disks are
typically of 1 Gb or greater storage capacity and more preferably
having at least 6 Gb of storage capacity. For example, in the
preferred embodiment described above each node of the Alta cluster
comprises a hard drive. Specifically, the head node has a hard
drive with 6 Gb of storage capacity whereas each sibling node has a
hard drive with 9 Gb of storage capacity. Other external components
include user interface device 1005, which can be a monitor and a
keyboard together with a pointing device 1006 such as a "mouse" or
other graphical input device. Typically, the computer system is
also linked to a network link 1007, which can be, e.g., part of an
Ethernet link to other local computer systems, remote computer
systems, or wide area communication networks such as the Internet.
For example, each computer system in the preferred Alta cluster of
computers described above is connected via an NFS network. This
network link allows the computer systems in the cluster to share
data and processing tasks with one another.
[0164] Loaded into memory during operation of this system are
several software components, which are both standard in the art and
special to the instant invention. These software components
collectively cause the computer system to function according to the
methods of the invention. The software components are typically
stored on mass storage 1004. Software component 1010 represents an
operating system, which is responsible for managing the computer
system and its network interconnections. The operating system can
be, for example, of the Microsoft Windows.TM. family, such as
Windows 98, Window 95 or Windows NT. Alternatively, the operating
system can be a Macintosh operating system, a UNIX operating system
or the LINUX operating system. Software component 1011 represents
common languages and functions conveniently present in the system
to assist programs implementing the methods specific to the present
invention. Languages that can be used to program the analytic
methods of the invention include, for example, UNIX or LINUX shell
command languages such as C, and C++; PERL; FORTRAN; HTML; and
JAVA. The methods of the present invention can also be programmed
or modeled in mathematical software packages which allow symbolic
entry of equations and high-level specification of processing,
including specific algorithms to be used, thereby freeing a user of
the need to procedurally program individual equations and
algorithms. Such packages include, e.g., Matlab from Mathworks
(Natick, Mass.), Mathematica from Wolfram Research (Champaign,
Ill.) or S-Plus from Math Soft (Seattle, Wash.). Accordingly,
software component 1012 represents analytic methods of the present
invention as programmed in a procedural language or symbolic
package.
[0165] In a preferred embodiment, the computer system contains a
software component 1013 which may be software for predicting (i.e.,
calculating) scores for one or more of the properties or parameters
described in Section 5.1.2, above (e.g., base composition,
position, perfect-match binding energy and/or cross-hybridization
binding energy to name a few) according to the methods described
above. Software component 1013 can also contain additional
programs, such as RepeatMasker or BLAST, that can also be used in
the methods of the present invention to evaluate nucleotide probes,
along with appropriate databases of nucleotide sequences for use in
conjunction with such programs.
[0166] For example the candidate probe sequences may be entered
directly by a user (e.g., using a keyboard or some other input
device) or may be loaded, e.g., from one or more databases stored
on a hard drive, a CD-ROM or some other storage medium, or from
other computer systems, e.g., over a the Internet. Alternatively,
the computer system may contain additional software components
10014 for generating candidate probe sequences, e.g., by randomly
generating sequences of a specified length or, in embodiments
wherein candidate probes for a particular target (e.g., for a
particular gene or genes) are sought, by generating oligonucleotide
sequences, e.g., according to the "tiling" methods described
hereinabove, that are complementary to various regions of the
target sequences.
[0167] The software component 1013 of the computer system also
preferably accepts one or more parameters or ranges of parameters
for use in selecting nucleotide probes. Exemplary parameters that
can be accepted by the software component include: probe length,
maximum distance from the 3'-end or 5'-end of the target sequence,
the maximum and/or minimum allowable binding energy scores
(including perfect-match and/or cross-hybridization binding
energies), upper and/or lower limits of acceptable base composition
(e.g., G %, C %, A %, T %, G+C %, A-C %), the longest permissible
single base run, and the maximum number of base overlap allowed
among different probes. The software component can also accept
parameters, such as temperature, salt concentration and target
polynucleotide concentration, for use in calculating hybridization
binding energies. These values can be input, e.g., directly by a
user or, alternatively, can be read by the software component from
a file.
[0168] Next, the user can cause execution of analysis software to
calculate one or more parameters or properties for each of the
candidate probes. In particular, the software preferably calculates
one or more of the particular properties described, e.g., in
Section 5.1.2, above, according to the methods and formulas
described in that section. For example, and not by way of
limitation, the analysis software can cause the processor to
calculate, e.g., the predicted sensitivity and specificity of the
candidate probes (e.g., according to Equations 8-10, above) or,
more preferably, the perfect-match binding energy and the minmax
cross-hybridization binding energy for each probe, e.g., using the
nearest neighbor model and Equation 1, above. Parameters for use in
the nearest neighbor model, particularly the dimer binding energies
and initiation parameters, may be entered by a user or loaded,
e.g., from a database. Alternatively, the analysis software 1013
may also comprise algorithms for calculating such parameters, e.g.,
according to Equations 6 or 7 above, and using experimental
hybridization data (e.g., hybridization intensities of a test
polynucleotide sequence to candidate probe sequences).
[0169] Preferably a computer system of the invention also contains
an analytical software component 1015 for ranking and/or selecting
candidate probes, e.g., for use in screening or signature chips
according to the methods described in the above sections. In one
preferred embodiment, a computer system may accept data relating to
the predicted sensitivities, specificities and other properties
(e.g., binding energies, base compositions, etc.) of a plurality of
candidate probes and use the data to rank the candidate probes
according to the methods described above in Section 5.1.3. These
data can be entered directly by a user, loaded from a database or,
more preferably determined by a computer system of the invention
using an analytical software component 1013 described above.
[0170] For example, and not by way of limitation, a computer system
may first calculate the sensitivities and specificities of a
plurality of candidate probes and then, using the calculated
sensitivities and specificities, rank the candidate probes
according to the ranking and/or selection algorithms described
hereinabove. Alternatively, sensitivities and specificities of a
plurality of candidate probes can be calculated by a first computer
system, and the results of such calculations can then be
transferred, e.g., by a network connection, to a second computer
system which ranks and/or selects the candidate probes according to
the ranking and/or selection algorithms.
[0171] In another preferred embodiment, experimental data
describing the sensitivity and specificity of a plurality of
candidate probes, such as the experimental data depicted in FIGS. 5
or 6, may loaded into a computer, e.g., directly by a user or from
a database, and the probes may be ranked according to the
above-described ranking and/or selection algorithms.
[0172] In a particularly preferred embodiment, the analytical
programs of a computer system of the invention cause the processor
element to execute the steps of the method depicted in FIG. 19 and
described, in detail, in Section 5.1.3, above.
[0173] The analytical systems of the invention also include
computer program products that contain one or more of the
above-described software components such that the software
components may be loaded into the main memory of a computer system.
Specifically, a computer program product of the invention includes
a computer readable storage medium having one or more computer
program mechanisms embedded or encoded thereon in a computer
readable format. The computer program mechanisms encode, e.g., one
or more of the analytical software components described above which
can be loaded into the memory of a computer system 1001 and cause
the processor of the computer system to execute the analytical
methods of the present invention.
[0174] The computer program mechanism or mechanisms are preferably
stored or encoded on a computer readable storage medium. Exemplary
computer readable storage media are discussed above and include,
but are not limited to: a hard drive, which may be, e.g., an
external or an internal hard drive of a computer system of the
invention, or a removable hard drive; a floppy disk; a CD-ROM; or a
tape such as a DAT tape. Other computer readable storage media will
also be apparent to those skilled in the art that can be used in
the computer program mechanisms of the present invention.
[0175] Alternative systems and methods for implementing the
analytic methods of this invention are intended to be comprehended
within the accompanying claims. In particular, the accompanying
claims are intended to include alternative program structures for
implementing the methods of this invention that will be readily
apparent to one of skill in the art.
5.3. MEASUREMENT OF HYBRIDIZATION LEVELS
[0176] In general, the hybridization methods of the present
invention can be performed using any probe or probes which comprise
a polynucleotide sequence and which are immobilized to a solid
support or surface. For example, as described supra, the probes may
comprise DNA sequences, RNA sequences, or copolymer sequences of
DNA and RNA. The polynucleotide sequences of the probes may also
comprise DNA and/or RNA analogues, or combinations thereof. For
example, the polynucleotide sequences of the probes may be full or
partial sequences of genomic DNA, cDNA, or mRNA sequences extracted
from cells. The polynucleotide sequences of the probes may also be
synthesized nucleotide sequences, such as synthetic oligonucleotide
sequences. The probe sequences can be synthesized either
enzymatically in vivo, enzymatically in vitro (e.g., by PCR), or
non-enzymatically in vitro.
[0177] The probe or probes used in the methods of the invention are
preferably immobilized to a solid support which may be either
porous or non-porous. For example, the probes of the invention may
be polynucleotide sequences which are attached to a nitrocellulose
or nylon membrane or filter. Such hybridization probes are well
known in the art (see, e.g., Sambrook et al., Eds., 1989, Molecular
Cloning: A Laboratory Manual, 2nd ed., Vol. 1-3, Cold Spring Harbor
Laboratory, Cold Spring Harbor, N.Y.). Alternatively, the solid
support or surface may be a glass or plastic surface.
5.3.1. MICROARRAYS GENERALLY
[0178] In a particularly preferred embodiment, hybridization levels
are measured to microarrays of probes consisting of a solid phase
on the surface of which are immobilized a population of
polynucleotides, such as a population of DNA or DNA mimics, or,
alternatively, a population of RNA or RNA mimics. The solid phase
may be a nonporous or, optionally, a porous material such as a gel.
Microarrays can be employed, e.g., for analyzing the
transcriptional state of a cell, such as the transcriptional states
of cells exposed to graded levels of a drug of interest, or to
graded perturbations to a biological pathway of interest.
[0179] In preferred embodiments, a microarray comprises a support
or surface with an ordered array of binding (e.g., hybridization)
sites or "probes" for products of many of the genes in the genome
of a cell or organism, preferably most or almost all of the genes.
Preferably the microarrays are addressable arrays, preferably
positionally addressable arrays. More specifically, each probe of
the array is preferably located at a known, predetermined position
on the solid support such that the identity (i.e., the sequence) of
each probe can be determined from its position in the array (i.e.,
on the support or surface). In preferred embodiments, each probe is
covalently attached to the solid support at a single site.
[0180] Microarrays can be made in a number of ways, of which
several are described below. However produced, microarrays share
certain characteristics: The arrays are reproducible, allowing
multiple copies of a given array to be produced and easily compared
with each other. Preferably, microarrays are made from materials
that are stable under binding (e.g., nucleic acid hybridization)
conditions. The microarrays are preferably small, e.g., between 5
cm.sup.2 and 25 cm.sup.2, preferably between 12 cm.sup.2 and 13
cm.sup.2. However, larger arrays are also contemplated and may be
preferable, e.g., for use in screening and/or signature chips
comprising a very large number of distinct oligonucleotide probe
sequences. Preferably, a given binding site or unique set of
binding sites in the microarray will specifically bind (e.g.,
hybridize) to the product of a single gene in a cell (e.g., to a
specific mRNA, or to a specific cDNA derived therefrom). However,
as discussed supra, in general other, related or similar sequences
will cross hybridize to a given binding site. Although there may be
more than one physical binding site per specific RNA or DNA, for
the sake of clarity the discussion below will assume that there is
a single, completely complementary binding site.
[0181] The microarrays of the present invention include one or more
test probes, each of which has a polynucleotide sequence that is
complementary to a subsequence of RNA or DNA to be detected. Each
probe preferably has a different nucleic acid sequence, and the
position of each probe on the solid surface is preferably known.
Indeed, the microarrays are preferably addressable arrays, and more
preferably are positionally addressable arrays. Specifically, each
probe of the array is preferably located at a known, predetermined
position on the solid support such that the identity (i.e., the
sequence) of each probe can be determined from its position on the
array (i.e., on the support or surface).
[0182] Preferably, the density of probes on a microarray is about
100 different probes (i.e., probes of non-identical sequence) per 1
cm.sup.2 or higher. More preferably, a microarray of the invention
will have at least 550 different probes per 1 cm.sup.2, at least
1,000 different probes per 1 cm.sup.2, at least 1,500 different
probes per 1 cm.sup.2 or at least 2,000 different probes per 1
cm.sup.2. In a particularly preferred embodiment, the microarray is
a high density array, preferably having a density of at least about
2,500 different probes per 1 cm.sup.2. The microarrays of the
invention therefore preferably contain probes of at least 2,500, at
least 5,000, at least 10,000, at least 15,000, at least 20,000, at
least 25,000, at least 50,000, at least 55,000, at least 100,000 or
at least 150,000 different (i.e., non-identical) sequences.
[0183] In one embodiment, the microarray is an array (i.e., a
matrix) in which each position represents a discrete binding site
for a product encoded by a gene (i.e., an mRNA or a cDNA derived
therefrom), and in which binding sites are present for products of
most or almost all of the genes in the organism's genome. For
example, the binding site can be a DNA or DNA analogue to which a
particular RNA can specifically hybridize. The DNA or DNA analogue
can be, e.g., a synthetic oligomer, a full-length cDNA, a less-than
full length cDNA, or a gene fragment.
[0184] Although in a preferred embodiment the microarray contains
binding sites for products of all or almost all genes in the target
organism's genome, such comprehensiveness is not necessarily
required. Usually the microarray will have binding sites
corresponding to at least about 50% of the genes in the genome,
often to at about 75%, more often to at least about 85%, even more
often to about 90%, and still more often to at least about 99%.
Alternatively, however, "picoarrays" may also be used. Such arrays
are microarrays which contain binding sites for products of only a
limited number of genes in the target organism's genome. Generally,
a picoarray contains binding sites corresponding to fewer than
about 50% of the genes in the genome of an organism.
[0185] Preferably, the microarray has binding sites for genes
relevant to the action of a drug of interest or in a biological
pathway of interest. A "gene" is identified as an open reading
frame (ORF) which encodes a sequence of preferably at least 50, 75,
or 99 amino acids from which a messenger RNA is transcribed in the
organism or in some cell in a multicellular organism. The number of
genes in a genome can be estimated from the number of mRNAs
expressed by the organism, or by extrapolation from a well
characterized portion of the genome. When the genome of the
organism of interest has been sequenced, the number of ORF's can be
determined and mRNA coding regions identified by analysis of the
DNA sequence. For example, the genome of Saccharomyces cerevisiae
has been completely sequenced, and is reported to have
approximately 6275 ORFs longer than 99 amino acids. Analysis of
these ORFs indicates that there are 5885 ORFs that are likely to
encode protein products (Goffeau et al., 1996, Science
274:546-567). In contrast, the human genome is estimated to contain
approximately 10.sup.5 genes.
5.3.2. PREPARING PROBES FOR MICROARRAYS
[0186] As noted above, the "probe" to which a particular
polynucleotide molecules specifically hybridizes according to the
invention is a complementary polynucleotide sequence. In one
embodiment, the probes of the microarray comprise nucleotide
sequences greater than about 250 bases in length corresponding to
one or more genes or gene fragments. For example, the probes may
comprise DNA or DNA "mimics" (e.g., derivatives and analogues)
corresponding to at least a portion of each gene in an organism's
genome. In another embodiment, the probes of the microarray are
complementary RNA or RNA mimics. DNA mimics are polymers composed
of subunits capable of specific, Watson-Crick-like hybridization
with DNA, or of specific hybridization with RNA. The nucleic acids
can be modified at the base moiety, at the sugar moiety, or at the
phosphate backbone. Exemplary DNA mimics include, e.g.,
phosphorothioates. DNA can be obtained, e.g., by polymerase chain
reaction (PCR) amplification of gene segments from genomic DNA,
cDNA (e.g., by RT-PCR), or cloned sequences. PCR primers are
preferably chosen based on known sequence of the genes or cDNA that
result in amplification of unique fragments (i.e., fragments that
do not share more than 10 bases of contiguous identical sequence
with any other fragment on the microarray). Computer programs that
are well known in the art are useful in the design of primers with
the required specificity and optimal amplification properties, such
as Oligo version 5.0 (National Biosciences). Typically each probe
on the microarray will be between 20 bases and 50,000 bases, and
usually between 300 bases and 1000 bases in length. PCR methods are
well known in the art, and are described, for example, in Innis et
al., eds., 1990, PCR Protocols: A Guide to Methods and
Applications, Academic Press Inc., San Diego, Calif. It will be
apparent to one skilled in the art that controlled robotic systems
are useful for isolating and amplifying nucleic acids.
[0187] An alternative, preferred means for generating the
polynucleotide probes of the microarray is by synthesis of
synthetic polynucleotides or oligonucleotides, e.g., using
N-phosphonate or phosphoramidite chemistries (Froehler et al.,
1986, Nucleic Acid Res. 14:5399-5407; McBride et al., 1983,
Tetrahedron Lett. 24:246-248). Synthetic sequences are typically
between about 15 and about 500 bases in length, more typically
between about 20 and about 100 bases, most preferably between about
40 and about 70 bases in length. In some embodiments, synthetic
nucleic acids include non-natural bases, such as, but by no means
limited to, inosine. As noted above, nucleic acid analogues may be
used as binding sites for hybridization. An example of a suitable
nucleic acid analogue is peptide nucleic acid (see, e.g., Egholm et
al., 1993, Nature 363:566-568; U.S. Pat. No. 5,539,083).
[0188] In alternative embodiments, the hybridization sites (i.e.,
the probes) are made from plasmid or phage clones of genes, cDNAs
(e.g., expressed sequence tags), or inserts therefrom (Nguyen et
al., 1995, Genomics 29:207-209).
5.3.3. ATTACHING PROBES TO THE SOLID SURFACE
[0189] The probes are attached to a solid support or surface, which
may be made, e.g., from glass, plastic (e.g., polypropylene,
nylon), polyacrylamide, nitrocellulose, gel, or other porous or
nonporous material. A preferred method for attaching the nucleic
acids to a surface is by printing on glass plates, as is described
generally by Schena et al, 1995, Science 270:467-470. This method
is especially useful for preparing microarrays of cDNA (See also,
DeRisi et al, 1996, Nature Genetics 14:457-460; Shalon et al.,
1996, Genome Res. 6:639-645; and Schena et al., 1995, Proc. Natl.
Acad. Sci. U.S.A. 93:10539-11286).
[0190] A second preferred method for making microarrays is by
making high-density oligonucleotide arrays. Techniques are known
for producing arrays containing thousands of oligonucleotides
complementary to defined sequences, at defined locations on a
surface using photolithographic techniques for synthesis in situ
(see, Fodor et al., 1991, Science 251:767-773; Pease et al., 1994,
Proc. Natl. Acad. Sci. US.A. 91:5022-5026; Lockhart et al., 1996,
Nature Biotechnology 14:1675; U.S. Pat. Nos. 5,578,832; 5,556,752;
and 5,510,270) or other methods for rapid synthesis and deposition
of defined oligonucleotides (Blanchard et al., Biosensors &
Bioelectronics 11:687-690). When these methods are used,
oligonucleotides (e.g., 20-mers) of known sequence are synthesized
directly on a surface such as a derivatized glass slide. Usually,
the array produced is redundant, with several oligonucleotide
molecules per RNA. Oligonucleotide probes can be chosen to detect
alternatively spliced mRNAs.
[0191] Other methods for making microarrays, e.g., by masking
(Maskos and Southern, 1992, Nuc. Acids. Res. 20:1679-1684), may
also be used. In principle, and as noted supra, any type of array,
for example, dot blots on a nylon hybridization membrane (see
Sambrook et al., supra) could be used. However, as will be
recognized by those skilled in the art, very small arrays will
frequently be preferred because hybridization volumes will be
smaller.
[0192] In a particularly preferred embodiment, microarrays of the
invention are manufactured by means of an ink jet printing device
for oligonucleotide synthesis, e.g., using the methods and systems
described by Blanchard in International Patent Publication No. WO
98/41531, published Sep. 24, 1998; Blanchard et al., 1996,
Biosensors and Bioelectronics 11:687-690; Blanchard, 1998, in
Synthetic DNA Arrays in Genetic Engineering, Vol. 20, J. K. Setlow,
Ed., Plenum Press, New York at pages 111-123. Specifically, the
oligonucleotide probes in such microarrays are preferably
synthesized in arrays, e.g., on a glass slide, by serially
depositing individual nucleotide bases in "microdroplets" of a high
surface tension solvent such as propylene carbonate. The
microdroplets have small volumes (e.g., 100 pL or less, more
preferably 50 pL or less) and are separated from each other on the
microarray (e.g., by hydrophobic domains) to form circular surface
tension wells which define the locations of the array elements
(i.e., the different probes).
5.3.4. TARGET POLYNUCLEOTIDE MOLECULES
[0193] As described, supra , the polynucleotide molecules which may
be analyzed by the present invention may be from any source,
including naturally occurring nucleic acid molecules, as well as
synthetic nucleic acid molecules. In a preferred embodiment, the
polynucleotide molecules analyzed by the invention comprise RNA,
including, but by no means limited to, total cellular RNA,
poly(A).sup.+ messenger RNA (mRNA), fraction thereof, or RNA
transcribed from cDNA (i.e., cRNA; see, e.g., Linsley &
Schelter, U.S. Pat. No. 6,271,002). Methods for preparing total and
poly(A).sup.+ RNA are well known in the art, and are described
generally, e.g., in Sambrook et al., supra. In one embodiment, RNA
is extracted from cells of the various types of interest in this
invention using guanidinium thiocyanate lysis followed by CsCl
centrifugation (Chirgwin et al., 1979, Biochemistry 18:5294-5299).
In an alternative embodiment, which is preferred for S. cerevisiae,
RNA is extracted from cells using phenol and chloroform, as
described in Ausubel et al. (Ausubel et al., eds., 1989, Current
Protocols in Molecular Biology, Vol III, Green Publishing
Associates, Inc., John Wiley & Sons, Inc., New York, at pp.
13.12.1-13.12.5). Poly(A).sup.+ RNA can be selected, e.g., by
selection with oligo-dT cellulose or, alternatively, by oligo-dT
primed reverse transcription of total cellular RNA. Cells of
interest include, but are by no means limited to, wild-type cells,
drug-exposed wild-type cells, modified cells, diseased cells and,
in particular, cancer cells.
[0194] In one embodiment, RNA can be fragmented by methods known in
the art, e.g., by incubation with ZnCl.sub.2, to generate fragments
of RNA. In one embodiment, isolated mRNA can be converted to
antisense RNA synthesized by in vitro transcription of
double-stranded cDNA in the presence of labeled dNTPs (Lockhart et
al., 1996, Nature Biotechnology 14:1675).
[0195] In other embodiments, the polynucleotide molecules to be
analyzed may be DNA molecules such as fragmented genomic DNA, first
strand cDNA which is reverse transcribed from mRNA, or PCR products
of amplified mRNA or cDNA.
5.3.5. HYBRIDIZATION TO MICROARRAYS
[0196] As described supra, nucleic acid hybridization and wash
conditions are chosen so that the polynucleotide molecules to be
analyzed by the invention (referred to herein as the "target
polynucleotide molecules) specifically bind or specifically
hybridize to the complementary polynucleotide sequences of the
array, preferably to a specific array site, wherein its
complementary DNA is located.
[0197] Arrays containing double-stranded probe DNA situated thereon
are preferably subjected to denaturing conditions to render the DNA
single-stranded prior to contacting with the target polynucleotide
molecules. Arrays containing single-stranded probe DNA (e.g.,
synthetic oligodeoxyribonucleic acids) may need to be denatured
prior to contacting with the target polynucleotide molecules, e.g.,
to remove hairpins or dimers which form due to self complementary
sequences.
[0198] Optimal hybridization conditions will depend on the length
(e.g., oligomer versus polynucleotide greater than 200 bases) and
type (e.g., RNA, or DNA) of probe and target nucleic acids. General
parameters for specific (i.e., stringent) hybridization conditions
for nucleic acids are described in Sambrook et al., (supra), and in
Ausubel et al., 1987, Current Protocols in Molecular Biology,
Greene Publishing and Wiley-Interscience, New York. When the cDNA
microarrays of Schena et al. are used, typical hybridization
conditions are hybridization in 5.times.SSC plus 0.2% SDS at
65.degree. C. for four hours, followed by washes at 25.degree. C.
in low stringency wash buffer (1.times.SSC plus 0.2% SDS), followed
by 10 minutes at 25.degree. C. in higher stringency wash buffer
(0.1.times.SSC plus 0.2% SDS) (Shena et al., 1996, Proc. Natl.
Acad. Sci. U.S.A. 93:10614). Useful hybridization conditions are
also provided in, e.g., Tijessen, 1993, Hybridization With Nucleic
Acid Probes, Elsevier Science Publishers B. V. and Kricka, 1992,
Nonisotopic DNA Probe Techniques, Academic Press, San Diego,
Calif.
[0199] Particularly preferred hybridization conditions for use with
the screening and/or signaling chips of the present invention
include hybridization at a temperature at or near the mean melting
temperature of the probes (e.g., within 5.degree. C., more
preferably within 2.degree. C.) in 1 M NaCl, 50 mM MES buffer (pH
6.5), 0.5% sodium sarcosine and 30% formamide.
5.3.6 SIGNAL DETECTION AND DATA ANALYSIS
[0200] It will be appreciated that when cDNA complementary to the
RNA of a cell is made and hybridized to a microarray under suitable
hybridization conditions, the level of hybridization to the site in
the array corresponding to any particular gene will reflect the
prevalence in the cell of mRNA transcribed from that gene. For
example, when detectably labeled (e.g., with a fluorophore) cDNA
complementary to the total cellular mRNA is hybridized to a
microarray, the site on the array corresponding to a gene (i.e.,
capable of specifically binding the product of the gene) that is
not transcribed in the cell will have little or no signal (e.g.,
fluorescent signal), and a gene for which the encoded mRNA is
prevalent will have a relatively strong signal.
[0201] In preferred embodiments, cDNAs from two different cells are
hybridized to the binding sites of the microarray. In the case of
drug responses, one cell is exposed to a drug and another cell of
the same type is not exposed to the drug. The cDNA derived from
each of the two cell types are differently labeled so that they can
be distinguished. In one embodiment, for example, cDNA from a cell
treated with a drug is synthesized using a fluorescein-labeled
dNTP, and cDNA from a second cell, not drug-exposed, is synthesized
using a rhodamine-labeled dNTP. When the two cDNAs are mixed and
hybridized to the microarray, the relative intensity of signal from
each cDNA set is determined for each site on the array, and any
relative difference in abundance of a particular mRNA is thereby
detected.
[0202] In the example described above, the cDNA from the
drug-treated cell will fluoresce green when the fluorophore is
stimulated, and the cDNA from the untreated cell will fluoresce
red. As a result, when the drug treatment has no effect, either
directly or indirectly, on the relative abundance of a particular
mRNA in a cell, the mRNA will be equally prevalent in both cells,
and, upon reverse transcription, red-labeled and green-labeled cDNA
will be equally prevalent. When hybridized to the microarray, the
binding site(s) for that species of RNA will emit wavelength
characteristic of both fluorophores. In contrast, when the
drug-exposed cell is treated with a drug that, directly or
indirectly, increases the prevalence of the mRNA in the cell, the
ratio of green to red fluorescence will increase. When the drug
decreases the mRNA prevalence, the ratio will decrease.
[0203] The use of a two-color fluorescence labeling and detection
scheme to define alterations in gene expression has been described,
e.g., in Shena et al., 1995, Science 270:467-470. An advantage of
using cDNA labeled with two different fluorophores is that a direct
and internally controlled comparison of the mRNA levels
corresponding to each arrayed gene in two cell states can be made,
and variations due to minor differences in experimental conditions
(e.g., hybridization conditions) will not affect subsequent
analyses. However, it will be recognized that it is also possible
to use cDNA from a single cell, and compare, for example, the
absolute amount of a particular mRNA in, e.g., a drug-treated or
pathway-perturbed cell and an untreated cell.
[0204] When fluorescently labeled probes are used, the fluorescence
emissions at each site of a transcript array can be, preferably,
detected by scanning confocal laser microscopy. In one embodiment,
a separate scan, using the appropriate excitation line, is carried
out for each of the two fluorophores used. Alternatively, a laser
can be used that allows simultaneous specimen illumination at
wavelengths specific to the two fluorophores and emissions from the
two fluorophores can be analyzed simultaneously (see Shalon et al.,
1996, Genome Res. 6:639-645). In a preferred embodiment, the arrays
are scanned with a laser fluorescent scanner with a computer
controlled X-Y stage and a microscope objective. Sequential
excitation of the two fluorophores is achieved with a multi-line,
mixed gas laser, and the emitted light is split by wavelength and
detected with two photomultiplier tubes. Such fluorescence laser
scanning devices are described, e.g., in Schena et al., 1996,
Genome Res. 6:639-645. Alternatively, the fiber-optic bundle
described by Ferguson et al., 1996, Nature Biotech. 14:1681-1684,
may be used to monitor mRNA abundance levels at a large number of
sites simultaneously.
[0205] Signals are recorded and, in a preferred embodiment,
analyzed by computer, e.g., using a 12 bit analog to digital board.
In one embodiment, the scanned image is despeckled using a graphics
program (e.g., Hijaak Graphics Suite) and then analyzed using an
image gridding program that creates a spreadsheet of the average
hybridization at each wavelength at each site. If necessary, an
experimentally determined correction for "cross talk" (or overlap)
between the channels for the two fluors may be made. For any
particular hybridization site on the transcript array, a ratio of
the emission of the two fluorophores can be calculated. The ratio
is independent of the absolute expression level of the cognate
gene, but is useful for genes whose expression is significantly
modulated by drug administration, gene deletion, or any other
tested event.
[0206] According to the method of the invention, the relative
abundance of an mRNA in two cells or cell lines is scored as
perturbed (i.e., the abundance is different in the two sources of
mRNA tested) or as not perturbed (i.e., the relative abundance is
the same). As used herein, a difference between the two sources of
RNA of at least a factor of about 25% (i.e., RNA is 25% more
abundant in one source than in the other source), more usually
about 50%, even more often by a factor of about 2 (i.e., twice as
abundant), 3 (three times as abundant), or 5 (five times as
abundant) is scored as a perturbation. Present detection methods
allow reliable detection of difference of an order of about 3-fold
to about 5-fold, but more sensitive methods are expected to be
developed.
[0207] It is, however, also advantageous to determine the magnitude
of the relative difference in abundances for an mRNA in two cells
or in two cell lines. This can be carried out, as noted above, by
calculating the ratio of the emission of the two fluorophores used
for differential labeling, or by analogous methods that will be
readily apparent to those of skill in the art.
6. EXAMPLE
[0208] The following examples are presented as exemplary
illustrations of the methods and compositions described hereinabove
and are not limiting of that description in any way. In particular,
the example presented in Section 6.1, below, describes particular
screening chips as well as their use to identify changes in mRNA
transcripts in unactivated and activated human lymphocytes,
respectively. A comparison of this data with conventional "spotter
chips" is also disclosed. Section 6.1 also discloses exemplary
signature chips and their use to further analyze changes in
signature genes identified using the screening and spotter chips.
These data verify that, although the screening chips of the
invention may occasionally fail to identify some changes in gene
expression, positive results obtained with such chips are
indicative of significant changes in gene expression. The chips are
therefore useful for screening large numbers of genetic transcripts
for changes in expression.
[0209] The example presented in Section 6.2 demonstrates the
effects of certain other probe design parameters on the reporting
properties of candidate oligonucleotide probes. Specifically, the
example demonstrates the effects which base composition,
information-content and the position of a candidate probe sequence
within a target gene sequence have or may have on the reporting
properties of a candidate oligonucleotide probe. Thus, these
properties are also useful for ranking and/or selecting probes for
use, e.g., in the screening chips and signature chips of the
present invention.
6.1. SYNTHESIS AND TESTING OF SCREENING CHIPS AND SIGNATURE
CHIPS
[0210] Two different types of microarrays or "chips" were used to
screen mRNA samples. The first microarray was a screening chip
comprising approximately 6,000 different polynucleotide probes that
were each 60 bases in length. The polynucleotide sequence of each
probe was selected according to the methods described above so that
each probe would hybridize sensitively and specifically to a
different gene transcript. The probes were synthesized in
microarrays using inkjet printing techniques described by Blanchard
(see, e.g., International Patent Publication No. WO 98/41531,
published Sep. 24, 1998; Blanchard et al., 1996, Biosensors and
Bioelectronics 11:687-690; Blanchard, 1998, in Synthetic DNA Arrays
in Genetic Engineering, Vol. 20, J. K. Setlow, Ed., Plenum Press,
New York at pages 111-123).
[0211] The second microarray, which is referred to herein as a
"spotter" chip, was a microarray of probes obtained from fragmented
double-stranded cDNA sequences of the same gene transcripts.
[0212] Aliquots of the same two RNA samples were hybridized to both
types of chips. Specifically, the RNA samples were samples of RNA
from unactivated and activated human lymphocyte cells from which
mRNA was amplified by PCR, using a primer containing a T7 RNA
polymerase promoter, and subsequently transcribed into labeled cRNA
(see, U.S. Pat. No. 6,271,002). A total of 550 genes was identified
that changed significantly (P-value<0.01) between the two
samples. Significant changes (P-value<0.01) were observed for
164 of these genes on both the screening and spotter chips.
However, 237 genes showed significant changes (P-value<0.01)
only on the screening chips, and 149 showed significant changes
(P-value<0.01) only on the spotter chips.
[0213] Signature chips were prepared to examine the behavior of
these 550 "signature genes" in more detail. Specifically, the
signature chips had, on average, 17 60mer polynucleotide probe
sequences for each of the signature genes (standard deviation
.+-.7) which were selected according to the above-described methods
of the invention so that each probe would hybridize sensitively and
specifically to one of the 550 signature genes. The selected
polynucleotide sequences were printed twice on a microarray using
the same inkjet printing methods, and the microarrays were
hybridized with the same RNA samples as the screening and spotter
chips.
[0214] FIGS. 14A-C compare the results obtained using the screening
and signature chips. Specifically, the figures each show scatter
plots of the change in each signature gene from unactivated to
activated lymphocytes measured with the signature chip (horizontal
axis) and screening chip (vertical axis). FIG. 14A compares the
results for the 164 genes that had significant changes on both the
screening and spotter chips. The correlation coefficient for the
measurements for these particular signature genes obtained using
the screening and signature chips was 0.97. FIG. 14B is an
identical scatter plot of the 237 genes for which significant
changes were observed with the screening chips but not with the
spotter chips. The correlation coefficient for the data in this
plot is 0.93. Thus, signature genes identified using a single
"best" 60 mer polynucleotide probe to detect each gene (i.e., on a
screening chip) were verified using the signature chips with
multiple 60 mer polynucleotide probes for each gene. In particular,
of the 401 signature genes that were identified on the screening
chip, 383 or .about.96% were reproducibly detected using the
signature chip. Thus, the false positive detection rate of the
screening chips was .gtoreq.4%.
[0215] A scatter plot comparing the results for the remaining 149
signature genes is shown in FIG. 14C. These signature genes showed
significant changes from unactivated to activated lymphocytes on
the spotter chips, but not on the screening chips. The correlation
coefficient for the changes in expression of these genes measured
by the signature and screening chips is only 0.69: Thus, regulation
of some genes may not be detected using a single oligonucleotide
probe for each gene although it can be detected using multiple
oligonucleotide probes per gene. The data from these 149
"false-negatives" was examined more closely by constructing
signature plots for each of the 149 signature genes. Specifically,
the signature plots compared the log of average intensity of the
hybridization signal measured with each probe on the signature chip
(horizontal axis) to the log ratio of hybridization intensity
between activated and unactivated cells (vertical axis). Exemplary
signature plots for four of the 149 gene are shown in FIGS. 15A-D.
In each plot, the corresponding probe or probes used on the
screening chips are indicated by open circles. In certain cases,
however, the exact oligonucleotide used as a probe on the screening
chip was not included on the signature chip. Accordingly, open
diamonds (e.g., in FIG. 15A) indicate oligonucleotide probes on the
signature chip that "bracketed" the probes used on the screening
chip and differed from that probe sequence by no more than five
bases. The log ratio of the measured signal from the screening
chips is indicated in FIGS. 15A-D by a solid line, whereas the log
ratio of measured signal from the spotter chips is indicated by a
dashed line.
[0216] The 149 "false-negative" genes could be generally divided
into four classes, with some genes being categorized in more than
one class. The first class, depicted in FIG. 15A, was characterized
by poor performance of the probe on the screening chip. In
particular, although most of the polynucleotide probes on the
signature chip exhibited results consistent with those of the
spotter chip, the particular probes used for the screening chip did
not yield good agreement with the spotter chip. Eighteen of the 149
false-negatives were categorized in this class.
[0217] The second class, depicted in FIG. 15B, was characterized by
threshold effects. In more detail, the polynucleotide probes used
in screening chips to detect genes in this class yielded results
that were consistent with results from the polynucleotide probes on
the signature chip. However, the fact that these genes are
apparently expressed at lower levels made changes in their
expression more difficult to detect using 60 mer polynucleotide
probes. Sixty-six of the false-negative genes belonged to this
class.
[0218] Polynucleotide probes for genes categorized in the third
class (FIG. 15C) did not give a clear consensus for why they gave
different results than the spotter chips. Most of the 39 genes
categorized in this class also exhibited weak regulation (i.e.,
<2-fold regulation) on the spotter chips, and therefore may have
been misreported by those chips. Finally, for most of the genes
categorized in class four (FIG. 15D), the 60 mer polynucleotide
probes on the signature and screening chips gave different results
than probes on the spotter chips. The probes on the spotter chips
may have been affected by cross-hybridization and therefore gave
inaccurate or misleading results. Thirty genes were categorized in
this class.
[0219] In conclusion, therefore, of the 149 signature genes that
were identified by the spotter chips but not by the screening
chips, only 84 of these appear to be true signature genes. Thus the
false-negative detection rate of the screening chip appears to have
only been about 15% (i.e., 84 false-negatives out of 550 signature
genes). 18 of these false-negatives, or .about.3%, were apparently
caused by selecting the wrong oligonucleotide probe for the
screening chip, whereas 66, or 12%, were because of variable
detection or threshold effects.
[0220] The data thus demonstrates that the screening chips of this
invention can be used, e.g., to screen large numbers of genetic
transcripts for changes in expression as a result of a change or
perturbation to a cell or organism. The number of both
false-positive and false-negative detections are reasonably low,
however, the higher rate of false-negative detection suggests that
the chips are most preferably used to screen for changes among many
transcripts since changes identified by such chips will most likely
be significant whereas a failure to detect a change is less
certain.
6.2. TESTING OF VARIOUS PROBE DESIGN PARAMETERS
[0221] This example describes methods and compositions which can be
used to assay the effects of various, exemplary oligonucleotide
probe parameters on their reporting properties. In particular, the
example describes the effects that oligonucleotide base
composition, the position of an oligonucleotide probe sequence
within a target nucleotide sequence, and sequence information
content have on the ability of candidate oligonucleotide probes to
reliably detect differential gene expression. The example thus
demonstrates that such parameters can be used, e.g., to rank and/or
screen candidate probes.
6.2.1. MATERIALS AND METHODS
[0222] Nucleotide sequences representing a plurality of human genes
from the NIH UniGene Collection (Available Web Site:
http://www.ncbi.nlm.nih.gov/UniGene; see also NCBI News, August
1996, Available Web Site:
http://www.ncbi.nlm.nkh.gov/Web/Newsltr/aug96.html; Schuler, 1997,
J. Mol. Med. 75:694-698; Schuler et al., 1996, Science 274:540-546;
Boguski & Schuler, 1995, Nature Genetics 10:369-371) were used
as sources of oligonucleotide probe sequences for experimental
validation of probe design parameters. Multiple 60-mer
oligonucleotides were then printed on inkjet chips for each target
gene sequence according to the methods described in Section 5.3.3,
above, and by Blanchard in International Patent Publication No. WO
98/41531, published Sep. 24, 1998 (see also Blanchard et al., 1996,
Biosensors and Bioelectronics 11:687-690; and Blanchard, 1998, in
Synthetic DNA Arrays in Genetic Engineering, Vol. 20, J. K. Setlow,
Ed., Plenum Press, New York at pages 111-123 for descriptions of
these inkjet printing methods)
[0223] Hybridization samples of cRNA were prepared for each assay
from total RNA extracted from cells using the method of Linsley and
Schelter (U.S. Pat. No. 6,271,002), and these samples were
hybridized to the microarray using conditions and methods described
in Section 5.3.5, below. In each assay hybridization levels were
measured for each probe using cRNA samples from perturbed and
unperturbed cells and the expression ratio of the target gene
reported by each probe was evaluated. The expression ratio reported
by each oligonucleotide probe was compared to the expected
expression change (i.e., the expected expression ratio) of the
target gene which had been previously determined, either from cDNA
microarray analysis, from the mean behavior of all oligonucleotide
probes derived from the target gene sequence or from the
literature.
6.2.2. BASE COMPOSITION
[0224] The percentage of guanine (G) and cytosine (C) bases in
oligonucleotide probes was compared to expression ratios reported
by each probe to evaluate how this parameter (i.e., base
composition) effected the probes' reporting properties.
Specifically, candidate oligonucleotide probes that are
complementary to the human transcription factor ETR103 (GenBank
Accession No. M62829) were designed and prepared on microarrays as
described hereinabove. cRNA samples from unactivated and activated
human lymphoblast cells were prepared and hybridized to the
microarrays and the ETR103 expression ration reported by each probe
was determined. ETR103 is known to be unexpressed in unactivated
lymphoblasts but is highly upregulated in activated lymphoblasts
(Shimizu et al., 1992, J. Biochem. 111:272-277).
[0225] FIG. 16 shows a plot of the ETR103 expression ratio
(unactivated lymphoblasts:activated lymphoblasts) reported for each
oligonucleotide probe (vertical axis) verses the fraction of G and
C bases in the probe (horizontal axis). Inspection of the plot
shows that there is a strong negative correlation between G and C
content in the oligonucleotide probes and the reliability with
which they reported the actual upregulation of ETR103 expression in
activated lymphoblasts. Only oligonucleotide probes with a G and C
content less than 0.4 (i.e. 40%) reported at least a two-fold
increase of ETR103 in activated lymphoblasts.
[0226] These data demonstrate, therefore, that the base content of
an oligonucleotide probe, such as the relative number of G and C
bases, is a parameter which can be used to screen candidate
oligonucleotide probes for a target gene. In particular, the data
demonstrates that it is preferable to select probes having a
relatively low number of G and C bases and, conversely, a
relatively high number of adenine (A) and thymine (T) bases.
6.2.3. TARGET SEQUENCE POSITION
[0227] Microarrays of oligonucleotide probes were also prepared to
evaluate how the distance of a probe's complementary sequence from
the 3' and 5' ends of its target gene sequence is related to the
probe's ability to detect changes in the expression of its target
gene. Specifically, "tiling chips" were prepared by selecting
oligonucleotide probes complementary to sequences within a target
gene that started at every third bases of the target gene. In one
particular experiment, described herein in detail, tiling chips
were prepared using oligonucleotide sequences complementary to
sequences of the target gene AML1b (GenBank Accession No. D43968).
cRNA samples were prepared from total RNA extracts from two
different human tissue culture cell lines: Jurkat and K562. These
cell lines are publicly available from the American Type Culture
Collection, (ATCC), 10801 25 University Boulevard, Manassa, Va.
20110-2209 (ATCC Accession Nos. TIB-152 and CCL-243,
respectively).
[0228] AML1b, which is specific to acute myeloid leukemia, is
expressed in K562 but not in Jurkat cells (Miyoshi et al., 1995,
Nucleic Acid Res. 23:2762-2769).
[0229] A plot of the expression ratio of AML1b in Jurkat cells:K562
cells reported by each probe (vertical axis) versus the distance of
the probe's complementary sequence from the 5' end of the AML1b
gene sequence (horizontal axis) is shown in FIG. 17. In general,
the reported expression ratios were constant until a threshold
distance from the 5' end of the AML1b gene sequence (approximately
4,000 nucleotides) was reached. Probes whose complementary
sequences were located at a distance less than about 4,000
nucleotides from the 5' end of the AML1b sequence were effectively
unable to detect any change of AML1b expression from Jurkat to K562
cells. This result is due to the fact that the cRNA hybridization
samples were prepared by initiating reverse transcription at the
3'-end of the expressed mRNA sequences. However, the generally
constant expression ratios shown in FIG. 17 for distances above
approximately 4,000 nucleotides from the 5' end are interrupted, at
intermittent intervals, by nonspecific oligonucleotide probes
within this distance that reported no expression ratio.
[0230] Thus, for hybridization samples prepared by reverse
transcription initiated at the 3' end of expressed mRNA sequences,
oligonucleotide probes used in the compositions and methods of the
invention should correspond to complementary sequences of the
target gene(s) that are within a certain threshold distance from
the 3' end of that target gene's sequence. Preferred, typical
threshold distances are generally within 5,000 bases of the 5' end
of a target gene sequence, and more preferably within 4,000 bases,
within 3,000 bases, within 2,000 bases or, most preferably, within
1,000 bases and still more preferably within 500 of the 3' end of a
target gene sequence.
[0231] Likewise, and as one skilled in the art readily appreciates,
in embodiments of the invention wherein hybridization samples are
prepared, e.g., from second strand synthesis initiated at the 5'
end of expressed mRNA sequences (for example by SMART RACE),
oligonucleotide probes used in the compositions and methods of the
invention should correspond to complementary sequences of the
target gene(s) that are within a certain threshold distance from
the 5' end of that target gene's sequence. Typically threshold
distances are generally within 5,000 bases of the 5' end of a
target gene sequence, and more preferably within 4,000 bases,
within 3,000 bases, within 2,000 bases or within 1,000 bases of the
5' end of a target gene sequence.
6.2.4. SEQUENCE INFORMATION CONTENT
[0232] Experiments were also performed to evaluate the correlation
between the "information content" of an oligonucleotide probe
sequence and the specificity with which that probe hybridizes to
its target gene sequence. Specifically, the program RepeatMasker
(Available Web Site:
http://ftp.genome.washington.edu/cgi-bin/RepeatMasker) was used to
identify low information-content sequences in target genes. Such
low information-content sequences consist of, e.g., simple repeats
of mono to hexanucleotide elements and complex elements found
repetitively in the genome. "Tiling chips" were prepared for these
target sequences, as described above, and the reported expression
ratios of the target genes in different cell types or cell lines
were evaluated.
[0233] Two particular experiments are described here in detail. In
the first experiment, chips were tiled with oligonucleotides
complementary to regions of the gene ETR103 (GenBank Accession No.
M62829) and were hybridized with cRNA samples prepared from total
RNA extracts of activated and unactivated lymphocytes as described
in Section 6.2.2, above. The ETR103 gene was also evaluated using
RepeatMasker to identify simple nucleotide repeat elements, such as
(CAG).sub.n, (CGG).sub.n and (AGGGGG).sub.n, within its
sequence.
[0234] In the second experiment, chips were tiled with
oligonucleotide complementary to regions of the gene AIM1 (GenBank
Accession No. U83115), a gene whose expression is associated with
the experimental reversal of tumorgenicity of human malignant
melanoma (Ray et al., 1997, Proc. Natl. Acad. Sci. US.A.
94:3229-3234) and is expressed in K562 cells but not in Jurkat
cells. The AIM1 gene was also evaluated using RepeatMasker, and an
ALU complex repetitive element was found within the transcribed
portion of this gene.
[0235] FIGS. 18A-B show the results from these two experiments. In
particular, FIG. 18A is a plot of the reported differential
hybridization (vertical axis) verses intensity (horizontal axis) of
oligonucleotide probes complementary to the ERT103 gene. Probes
that are complementary to regions of the ERT103 gene sequence that
were masked by RepeatMasker (i.e., regions containing the
repetitive element (CAG).sub.n, (CGG).sub.n or (AGGGGG).sub.n) are
indicated by open circles. FIG. 18B shows a plot of the reported
differential hybridization (vertical axis) verses intensity
(horizontal axis) of oligonucleotide probes complementary to the
AIM1 gene. Probes that are complementary to regions of the AIM1
gene for which greater the 60% of the sequence is contained within
the ALU repeat are indicated by open circles.
[0236] The results are surprisingly dramatic. FIG. 18A shows that
oligonucleotide probes that overlap with simple nucleotide repeats
are completely nonspecific and report no differential expression.
Likewise, as can be seen in FIG. 18B, overlap of oligonucleotide
probes with a complex repetitive element (e.g., the ALU repeat)
also decreases the specificity of the probe. However, more than
minimal overlap with such a complex repetitive element is required
for a complete loss in specificity. Probes for which greater than
60% of the oligonucleotide sequence overlaps with the repetitive
element do, however, exhibit a complete loss of specificity.
Smaller overlap of probe sequences with such elements result in
smaller decreases in the reported expression ratio. Thus, in the
probes selected for use in the methods and compositions of the
present invention, preferably less than 60% of a probe's sequence
overlaps with (i.e., is complementary to) repetitive elements such
as simple nucleotide repeats or complex repetitive elements. More
preferably, none (i.e., 0%) of a probe's sequence overlaps with
repetitive sequence elements.
7. REFERENCES CITED
[0237] All references cited herein are incorporated herein by
reference in their entirety and for all purposes to the same extent
as if each individual publication or patent or patent application
was specifically and individually indicated to be incorporated by
reference in its entirety for all purposes.
[0238] Many modifications and variations of this invention can be
made without departing from its spirit and scope, as will be
apparent to those skilled in the art. The specific embodiments
described herein are offered by way of example only, and the
invention is to be limited only by the terms of the appended claims
along with the full scope of equivalents to which such claims are
entitled.
* * * * *
References