Recursive descent probabilities for rare recessive lethals

Recursively computed descent probabilities provide an effective way to evaluate possible ancestries of rare alleles segregating in large and complex genealogies, but they ignore information other than the descent to a small set of current gene copies. We show how descent probability computations can be modified to incorporate the fact that no ancestor can be affected by a lethal recessive trait. This methodology is applied to two rare recessive traits segregating in a complex Hutterite (Lehrerleut) genealogy.