FastZ: accelerating gapped whole genome alignment on GPUs
FastZ: accelerating gapped whole genome alignment on GPUs
复制标题
FastZ:在 GPU 上加速有缺口的全基因组比对
DOI:
10.1145/3458817.3476202
复制
发表时间:
2021
期刊:
影响因子:
--
通讯作者:
Thottethodi, Mithuna
中科院分区:
文献类型:
--
作者:
Gundabolu, Sree Charan;Vijaykumar, T. N.;Thottethodi, Mithuna
Recognizing the importance of whole genome alignment (WGA), the National Institutes for Health maintains LASTZ, a sequential WGA application. As genomic data grows, there is a compelling need for scalable, high-performance WGA. Unfortunately, high-sensitivity, `gapped' alignment which uses dynamic programming (DP) is slow, whereas faster alignment with ungapped filtering is often less sensitive. We developFastZ, a GPU-accelerated, gapped WGA software which matches gapped LASTZ in sensitivity.FastZemploys a novelinspector-executorscheme in which (a) thelightweightinspector elides DP traceback except in common, extremely short alignments, where the inspector performs limited,eagertraceback to eliminate the executor, and (b)executor trimmingavoids unnecessary work. Further,FastZemploysregister-based cyclic-bufferingto drastically reduce memory traffic, andgroups DP problems by sizefor load balance.FastZrunning on an RTX 3080 GPU and our multicore implementation of LASTZ achieve 111x and 20x speedups over the sequential LASTZ, respectively.