DNA and biosequences are being determined faster thanthey can be compared and analyzed. Present techniques forrigorously searching a large database are costly and timeconsuming. Under all plausible growth projections, the problemwill soon become overwhelming. A custom, very large scaleintegrated circuit that could be used for high speed comparison,analysis, and interpretation of DNA and protein sequences has beendesigned. The systolic array can, in a single pass, scrutinizeeach segment of every biosequence in a database to determine itshomology to any segment of an input pattern according to astatistically rigorous criterion. Phase I research includesdesigning and prototyping a low cost biological sequencecomparative analysis node (BioSCAN) printed circuit board for anIBM-compatible AT personal computer. In addition, softwareenabling the board to interface with commonly used, public domainsoftware packages (as, e.g., FASTA or BLAST3) is being developed. Finally, the performance of the system in actual laboratorysettings is being evaluated against current techniques. In PhaseII, an interactive analysis platform will be developed on the basisof the evaluation results obtained in Phase I.Anticipated Results/Potential Commercial Applications as described by the awardee:The proposed BioSCAN system will scan a database ata sustained rate of 2.5 million sequence elements per second, orapproximately 1,000 times faster than the best software available. The result of a BioSCAN search may be useful by itself or may serveas prefiltered input for further analysis. A cost/performanceratio 10,000 times lower than that of a massively parallel orvector supercomputer is expected. The performance increaseafforded by this BioSCAN system will increase the productivity ofmolecular biologists, and the interactive capability will changetheir fundamental approach to the sequence analysis problem.
