Fig. 1

Our pipeline was identical for every sample, except for how we handled PCR duplicates. Original BAM files with mapped reads were processed using the shown pipeline. PCR duplicates were ignored, removed using SAMTools (rmdup), or marked (left in the file, but ignored in subsequent steps) using Picard (MarkDuplicates). After the duplicates step, all subsequent steps in the analysis pipeline were identical. The final output of the pipeline is a multi-sample VCF file. VCF: Variant Call Format, BQSR: Base Quality Score Recalibration, VQSR: Variant Quality Score Recalibration