Date of Publication :7th November 2016
Abstract: The problem of detecting similarities between new gene sequence and genomic sequences by search operation is fundamental to many research pursuits in bioinformatics. There are a number of software tools which take as input, a query string and a database string and return a list of approximate matches between the two, each with an associated alignment and numeric score. BLAST is a kind of software tool used to search genomic sequence. In BLAST the first stage is most time consuming and critical step. To reduce the time in this stage, filtration method is applied to reduce amount of computation data. This is done by using Bloom filter. A Hash function is associated with Bloom filter to generate the address where a particular data will be stored in the programming phase. In querying phase, the same Hash function has been utilized to check whether the data has been stored. There are different methods to implement Hash function. In this paper a new method to implement Hash function by using LFSR has been proposed. The generator polynomials used to implement LFSR is 32-bit and the usual serial implementation has been converted into a parallel using the CRC-32 code to speed-up the address generation step in programming and querying phase of the Bloom filter.
Reference :