Fig. 2

Benchmarking of the Phase 3 GIS parallelization HPC-GVCW as compare with the standard GATK pipeline using 30 resequenced rice accessions mapped to a single reference genome, a execution time and b CPU hours (execution time × number of nodes) for job completion. Notes: Comparisons were tested between the standard GATK pipeline without chunks using 1 node (blue dots), and HPC-GVCW using a range of computing nodes chunked length combinations, i.e., chunks sizes of 10 Kb, 100 Kb, 200 Kb, 500 Kb, 1 Mb, 5 Mb, 10 Mb, 20 Mb, and chromosome level, which use 2342, 237, 120, 50, 27, 8, 6, 5, and 4 nodes, respectively (yellow dots)