January 12-16, 2002
Town & Country Convention Center
San Diego, CA
Poster: General Comparative
Large-insert clone fingerprinting and contig assembly technology is being widely used to rapidly develop whole-genome, BAC-based physical maps of plants and animals. However, many significant questions remain to be investigated in order to construct high-quality physical maps efficiently using the BAC fingerprinting technology. We have investigated the relationship between BAC fingerprinting methods and accuracy of physical maps. First, we simulated fingerprints from random sequences using one, two, three and five restriction enzymes, followed by running on agarose gels, sequencing gels, 36-cm capillary sequencers, and 36-cm capillary sequencers, respectively, and then the fingerprints were assembled into overlapping contigs. Second, we selected a 2-Mb sequenced BAC contig of the human genome and a 2-Mb sequenced BAC contig of the Arabidopsis genome from the existing sequence database, created fingerprints from the contig BACs based on the restriction sites of their sequences using one, two, three and five enzymes separately, calculated the band mismatch percentage between the clones, and then assembled them into contigs. Third, we fingerprinted the same sets of contig BACs in laboratory, analyzed the fingerprints and assembled contigs as above. The relationship between accuracy of physical map assembly and fingerprinting methods is being analyzed. Preliminary results showed that using a 3-enzyme digestion with 2-dye labeling system and capillary sequencer detection is the best method for developing a whole-genome physical map from BACs. An average of 55 bands per BAC fingerprint resulted in optimal contig assembly with only 6% band mismatch.