Parallel Position Weight Matrices Algorithms
LIFL, UMR CNRS 8022, Universite Lille 1, INRIA Lille-Nord Europe, Lille, France
International Symposium on Parallel and Distributed Computing (ISPDC 2009) (2009) 65-69, Parallel Computing (26 October 2010)
@conference{giraud2009parallel,
title={Parallel position weight matrices algorithms},
author={Giraud, M. and Varre, J.S.},
booktitle={Parallel and Distributed Computing, 2009. ISPDC’09. Eighth International Symposium on},
pages={65–72},
year={2009},
organization={IEEE}
}
Position Weight Matrices (PWMs) are broadly used in computational biology. The basic problems, Scan and Multiscan, aim to find all the occurrences of a given PWM or a set of PWMs in long sequences. Some other PWM tasks share a common NP-hard subproblem, ScoreDistribution The existing algorithms rely on the enumeration on a large set of scores or words, and they are mostly not suitable for parallelization.We propose a new algorithm, BucketScoreDistribution, that is both very efficient and suitable for parallelization. We bound the error induced by this algorithm. We realized a GPU prototype for scan, multiscan and BucketScoreDistribution with the CUDA libraries, and report for the different problems speedups larger than 10x on several Nvidia cards.
November 22, 2010 by hgpu
