Parallel protein sequence matching on multicore computers

STRIKE was introduced and implemented to predict protein-protein interactions where proteins interact if they contain similar substrings of amino acids. On the yeast protein interaction literature, STRIKE was shown to improve upon the existing state-of-the-art methods for protein-protein interaction prediction. Herein, we describe the parallelization of STRIKE and its multithreaded implementation and performance enhancement on multicore systems. On large protein sequence sets, the execution time of a 16-thread implementation of this bioinformatics algorithm was reduced from about a week on a unithreaded implementation on a serial uniprocessor machine to 1.5 days on one quad core ×86 machine, down to 4.5 hours on 8 such quad core machines. Key optimizations to the implementation are also discussed.