PERFORMANCE STUDY OF PARALLEL HYBRID MULTIPLE PATTERN MATCHING ALGORITHMS FOR BIOLOGICAL SEQUENCES
C. Kouzinopoulos, Panagiotis D. Michailidis, Konstantinos G. Margaritis · 2012
Abstract: Multiple pattern matching is widely used in computational biology to locate any number of nucleotides in genome databases. Processing data of this size often requires more computing power than a sequential com-puter can provide. A viable and cost-effective solution that can offer the power required by computationally intensive applications at low cost is to share computational tasks among the processing nodes of a high per-formance hybrid distributed and shared memory platform that consists of cluster workstations and multi-core processors. This paper presents experimental results and a theoretical performance model of the hybrid im-plementations of the Commentz-Walter, Wu-Manber, Set Backward Oracle Matching and the Salmela-Tarhio-Kytöjoki family of multiple pattern matching algorithms when executed in parallel on biological sequence databases. 1