Using parallel algorithms for searching molecular sequence databases

Abstract

This work presents the development of algorithms for approximate string matching using parallel methods. It intends to do the maximum of molecular sequences comparisons per unity of time. The parallel program implementation has carried out in C on an available twenty processing nodes clustering architecture using a model of parallel programming systems, the MPI (Message-Passing Interface), which is as library of subroutines. We use one of the possible approaches to reduce the time spent on comparisons of molecular database sequences by distributing the data among processors, which achieves a linear speedup (time) and requires constant space memory per processor.