Shotgun sequencing of the human transcriptome with ORF expressed sequence tags

Theoretical considerations predict that amplification of expressedgene transcripts by reverse transcription–PCR using arbitrarilychosen primers will result in the preferential amplification of thecentral portion of the transcript. Systematic, high-throughputsequencing of such products would result in an expressed se-quence tag (EST) database consisting of central, generally codingregions of expressed genes. Such a database would add significantvalue to existing public EST databases, which consist mostly ofsequencesderivedfromtheextremitiesofcDNAs,andfacilitatetheconstruction of contigs of transcript sequences. We tested ourpredictions, creating a database of 10,000 sequences from humanbreast tumors. The data confirmed the central distribution of thesequences, the significant normalization of the sequence popula-tion, the frequent extension of contigs composed of existinghuman ESTs, and the identification of a series of potentiallyimportant homologues of known genes. This approach shouldmake a significant contribution to the early identification ofimportant human genes, the deciphering of the draft humangenome sequence currently being compiled, and the shotgunsequencing of the human transcriptome.