Performance and Memory Analysis of the Pa-Star Alignment Program on the Santos Dumont Supercomputer

• Kelen Souza LNCC / FAETERJ-RJ
• Micaella Coelho LNCC
• Carla Osthoff LNCC
• Kary Ocaña LNCC

DOI: https://doi.org/10.5753/eradrj.2024.4392

Abstract

This study evaluated the performance of the bioinformatics program PA-Star, used for multiple sequence alignment, on two architectures of the Santos Dumont supercomputer: the Mesca2, with higher computational and memory capacity, and the Intel Xeon Ivy Bridge node. The biological sequences processed varied in size, quantity, and level of similarity. It was observed that Mesca2 showed superior performance for files with high RAM demand, while the Intel Xeon Ivy Bridge was more efficient for files with lower demand. These results suggest that Mesca2 is essential for processing alignments with high RAM consumption on Santos Dumont.

References

Chowdhury B, Garai G (2017) A review on multiple sequence alignment from the perspective of genetic algorithm. Genomics 109:419–431. DOI: 10.1016/j.ygeno.2017.06.007

Sundfeld, D. Alinhamento primário e secundário de sequências biológicas em arquiteturas de alto desempenho. 2017, 167 f., il. Tese (Doutorado em Informática). Universidade de Brasília, 2017.

Sundfeld, D; Razzolini, C; Teodoro, G; Boukerche, A; Melo, ACM. PA-Star: A disk-assisted parallel A-Star strategy with locality-sensitive hash for multiple sequence alignment. Journal of Parallel and Distributed Computing, v. 112, p. 154-165, 2018.