The utilization of DNA sequencing data for diagnosis and drug discovery promises to be a revolutionary change in modern medicine; however, time to analysis can be a major roadblock. On average, a whole human genome generates 100 GB of compressed sequencing data which must be processed using nearly one thousand CPU-hours to generate conclusions of biological significance for geneticists, bioinformaticians and physicians.
Faster analysis will result in faster discoveries, reduced costs and enable genomic analysis to be part of standard healthcare practices. This whitepaper from Parabricks examines the compute-intensive nature of genomic analysis, outlines their accelerated workflow for alignment, pre-processing and variant calling, and demonstrates how GPUs can accelerate parallelizable sequencing tasks in genome processing at different stages to minimize total analytics time, meeting the genome sequencing processing time challenge.