A Novel Algorithm for In-silico Est Expression Profiling

Motivation: Various tasks in computational biology, including primer and oligonucleotide array design; epitope mapping, etc. require finding unique regions in gene sequences. We developed a novel algorithm that finds such unique regions based on an alignment of the gene sequence to its paralogs. This algorithm was utilized for in-silico expression profiling using EST databases – an accurate low cost alternative to high-throughput “wet bench” expression profiling methods. Results: Analysis of expression patterns of enzymes in glycolytic pathway led to alternative hypothesis of fructose metabolism in the brain. Publicly available database of Expressed Sequence Tags (dbEST) consisting of multiple flat files was mined for crucial information and reformatted into a relational database. A public Internet access to the database is planned. Availability: -Available as a commercial package through Boston University technology transfer. -Free availability over the Internet: watch for update at http://zlab.bu.edu/~dmitriyl.