Pilot Evaluation of Saliva Samples as An Alternative to Blood for Large-scale Whole-genome Sequencing Studies

Abstract

Whole-genome sequencing (WGS) studies have traditionally relied on blood as the primary DNA source and manual workflows. As population-based genomics studies expand, minimally invasive and easily deployable biospecimen collection methods compatible with high-throughput automation are increasingly needed. In this pilot study, we established a real-world, end-to-end automated WGS workflow, encompassing robotic DNA extraction, library preparation, sequencing and quality control. We compared saliva- and blood-derived WGS across key analytical parameters, including performance at each step of the workflow, overall cost, and feasibility for future large-scale implementation. First, germline variant calls from saliva and blood samples showed high concordance, reaching 97% genome-wide and over 99% in high-confidence genomic regions. Second, key workflow metrics, including DNA quality and sequencing quality, were comparable between saliva and blood samples, while saliva samples had a lower per-sample cost. We compared three commercially available saliva collection kits and found that all achieved at least 95% genome-wide concordance with matched blood samples, with differences primarily driven by cost. Finally, automated saliva processing produced sequencing and short variant calling metrics broadly comparable to manual processing in a small paired comparison. Collectively, these results provide a practical and scalable solution for incorporating saliva into future large cohort sequencing studies. Larger and more diverse studies are needed to further validate the robustness and generalizability of this pilot study.