The impact of DNA source and eukaryotic DNA enrichment on genetic variant detection from human whole-genome sequencing data

With sequencing costs continuing to drop, whole-genome sequencing (WGS) is being used increasingly frequently for research studies and clinical genetic diagnostics. While blood is the most common source of DNA for WGS, saliva and buccal samples are more stable, less invasive to collect, and can be self-collected. However, it is not known whether DNA from saliva and buccal samples can be genotyped as accurately as DNA from blood. Using matched blood, saliva, and buccal samples from four individuals, we show that DNA source has little effect on the accuracy of single nucleotide variant and indel detection, but that copy number variants (CNVs) can be detected more accurately from blood-derived DNA. We also show that methylation-based capture to enrich for eukaryotic DNA in saliva and buccal samples is effective in removing bacterial DNA, but also reduces read-depth uniformity and thus the accuracy of read depth-based CNV detection. Overall, we recommend using DNA extracted from blood rather than saliva or buccal swabs for WGS; if saliva or buccal samples are used, we recommend against using methylation-based eukaryotic DNA enrichment.