Fast and Accurate Genomic Analyses using Genome Graphs

Abstract The human reference genome serves as the foundation for genomics by providing a scaffold for alignment of sequencing reads, but currently only reflects a single consensus haplotype, which impairs read alignment and downstream analysis accuracy. Reference genome structures incorporating known genetic variation have been shown to improve the accuracy of genomic analyses, but have so far remained computationally prohibitive for routine large-scale use. Here we present a graph genome implementation that enables read alignment across 2,800 diploid genomes encompassing 12.6 million SNPs and 4.0 million indels. Our Graph Genome Pipeline requires 6.5 hours to process a 30x coverage WGS sample on a system with 36 CPU cores compared with 11 hours required by the GATK Best Practices pipeline. Using complementary benchmarking experiments based on real and simulated data, we show that using a graph genome reference improves read mapping sensitivity and produces a 0.5% increase in variant calling recall, or about 20,000 additional variants being detected per sample, while variant calling specificity is unaffected. Structural variations (SVs) incorporated into a graph genome can be genotyped accurately under a unified framework. Finally, we show that iterative augmentation of graph genomes yields incremental gains in variant calling accuracy. Our implementation is a significant advance towards fulfilling the promise of graph genomes to radically enhance the scalability and accuracy of genomic analyses.