Aleksandar Milosavljevic, Suzana Savkovic, Radomir Crkvenjakov, David Salbego, Hope Serrato, Heidi Kreuzer, Anne Gemmell, Shawna Batus, Danica Grujic, Susan Carnahan, and Jovanka Tepavcevic
Recently developed hybridization technology (Drmanac et al. 1994) enables economical large-scale detection of short ofigomers within DNA fragments. Tile newly developed recognition method (Milosavljevid 1995b) enables comparison of lists of oligomers detected within DNA fragments against known DNA sequences. We here describe an experiment involving a set of 4,513 distinct genomic E.coli clones of average length 2kb, each hybridized with 636 randomly selected short oligomer probes. High hybridization signal with a particular probe was used as an indication of the presence of a complementary oligomer in the particular clone. For each clone, a list of oligomers with highest hybridization signals was compiled. The database consisting of 4,513 oligomer lists was then searched using known E.coli sequences as queries in an attempt to identify the clones that match the query sequence. Out of a total of 11 clones that were recognized at highest significance level by our method, 8 were single-pass sequenced from both ends. The single-pass sequenced ends were then compared against the query sequences. The sequence comparisons confirmed 7 out of the total of 8 examined recognitions. This experiment represents the first successful example of genome-scale sequence recognition based on hybridization data.