Whole-Genome Sequencing Data
Whole-genome sequencing data for the family were downloaded from the Complete Genomics website. Family members were originally sequenced to an average genome-wide coverage of 80×. We used variants called by the Complete Genomics Analysis Pipeline (v.2.0.0). We performed an additional filtering step testing for Mendelian inconsistency to obtain a high-confidence set of variants, and we eventually retained 5,546,682 out of the original 6,181,281 SNPs. We further compared our selected variants to those assessed by long-fragment read (LFR) technology (N50s 400–1,500 kb).32 LFR has a claimed error rate of 1 in 10 Mb. Our comparison showed that variant concordance between the 80× shotgun-sequencing approach and LFR technology was 99.91% to 99.95% (Table S2). In addition, the same family was also sequenced to 50× by Illumina Platinum Genomes, and the genotyping concordance with Complete Genomics was found to be 99.62% to 99.83% (Table S4).