A Data Parallel Strategy for Aligning Multiple Biological Sequences on Homogeneous Multiprocessor Platform

Abstract

In this paper, we address the biological sequence alignment problem, which is a fundamental operation performed in computational biology. We employ the data parallelism paradigm that is suitable for handling large-scale processing to achieve a high degree of parallelism. Using data parallelism, we propose a strategy in which we employ a parallel clustering scheme to partition the set of sequences into subsets based on sequence similarity. Then the subsets are distributed among the processors using a heuristic algorithm based on Integer Programming so as to minimize the overall processing time, and each subset can be independently aligned in parallel using any sequential approach. The global alignment is achieved using a progressive profile-profile alignment within and between the processors. We implement the proposed algorithm on a cluster using the MPI library, and analyze the experimental results for different problem sizes in terms of quality of alignment, execution time and speed-up.