Fig. 1.

Analysis of child-other segments in parents. (A) The majority of child-other segments are not detected in either parent. Parent segment overlap is calculated as the percentage of sites in the child-other segment that are included in the parent-other segment. (B) Truncation points for parent-other segments are nearly always confidently genotyped opposite homozygote sites, consistent with false-positive IBD in the child. The opposite homozygote site causing truncation of the parent-other segment was examined in a randomly selected subset of all 3,371,616 segments with partial parent overlap. (C) Child-other segments with no corresponding parent-other segments contain many parent-other opposite homozygotes in the region, also consistent with false-positive IBD in the child. For each of these child-other segments, the number of opposite homozygote sites present between the parent and the other individual at that segment location is calculated separately for each parent, and the smaller is chosen as the number of opposite homozygotes in the region.