A Hybrid OpenMP-MPI Parallelization of Structure Software

Big Data has an increasing impact on the use of bioinformatics software. One way to deal with this challenge is through parallel computing. Using the program Structure as a case study, this pa-per investigates ways in which to counteract the challenges created by the growing datasets. This paper proposes an OpenMP-MPI hy-brid parallelization of the MCMC steps, which are an integral part of Structure, and analyses the performance under various scenar-ios. The results indicate that the parallelization produce significant speedups over the serial version in all scenarios tested. This al-lows for the use of the hardware in a more efficient manner, by adapting the program to the parallel architecture. This is important because not only does it reduce the time required to perform exist-ing analyses, but also opens the door to the analysis of previously impractically large datasets. Keywords