We created a panel of informative genotypes to confirm family relationships, and we omitted trios if expected family relationships did not confirm, or if there were unexpected relationships within or across families (Supplemental Experimental Procedures). Additionally, we removed samples with excess de novo variants (> 5).
