Mid-pass whole genome sequencing enables biomedical genetic studies of diverse populations
Anne‐Katrin Emde, Amanda Phipps‐Green, Murray Cadzow, Clair Gallagher, Tanya J. Major, Marilyn E. Merriman, Ruth Topless, Riku Takei, Nicola Dalbeth, Rinki Murphy, Lisa K. Stamp, Janak de Zoysa, Philip L. Wilcox, Keolu Fox, Kaja A. Wasik, Tony R. Merriman, Stephane E. Castel
Abstract
BACKGROUND: Historically, geneticists have relied on genotyping arrays and imputation to study human genetic variation. However, an underrepresentation of diverse populations has resulted in arrays that poorly capture global genetic variation, and a lack of reference panels. This has contributed to deepening global health disparities. Whole genome sequencing (WGS) better captures genetic variation but remains prohibitively expensive. Thus, we explored WGS at "mid-pass" 1-7x coverage. RESULTS: Here, we developed and benchmarked methods for mid-pass sequencing. When applied to a population without an existing genomic reference panel, 4x mid-pass performed consistently well across ethnicities, with high recall (98%) and precision (97.5%). CONCLUSION: Compared to array data imputed into 1000 Genomes, mid-pass performed better across all metrics and identified novel population-specific variants with potential disease relevance. We hope our work will reduce financial barriers for geneticists from underrepresented populations to characterize their genomes prior to biomedical genetic applications.