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.