LARGE-SCALE CONCATENATION CDNA SEQUENCING

A total of 100 kb of DNA derived from 69 individual human brain cDNA c lones of 0.7-2.0 kb were sequenced by concatenated cDNA sequencing (CC S), whereby multiple individual DNA fragments are sequenced simultaneo usly in a single shotgun library. The method yielded accurate sequence s and a similar efficiency compared with other shotgun libraries const ructed from single DNA fragments (>20 kb). Computer analyses were carr ied out on 65 cDNA clone sequences and their corresponding end sequenc es to examine both nucleic acid and amino acid sequence similarities i n the databases. Thirty-seven clones revealed no DNA database matches, 12 clones generated exact matches (greater than or equal to 98% ident ity), and 16 clones generated nonexact matches (57%-97% identity) to e ither known human or other species genes. Of those 28 matched clones, 8 had corresponding end sequences that failed to identify similarities . In a protein similarity search, 27 clone sequences displayed signifi cant matches, whereas only 20 of the end sequences had matches to know n protein sequences. Our data indicate that full-length cDNA insert se quences provide significantly more nucleic acid and protein sequence s imilarity matches than expressed sequence tags (ESTs) for database sea rching.