This analysis sought to determine the impact of specific ascertainment crit
eria based upon nuclear family affectation structures. Specifically, we eva
luated the predicted and observed proportion of alleles shared identical by
descent conditional on the number of affected and unaffected siblings in a
pedigree, and compared sib-pair method linkage results under two ascertain
ment schemes, random vs. selected ascertainment, for this simulated complex
genetic disease. These results suggest that samples differing in the compo
sition of affected and unaffected siblings in the family will differ in the
ir power to detect linkage. An effect of sampling scheme on power to map us
ing affected-sib-pair methods should be considered when a reported linkage
is not found in another study population. ((C)) 1999 Wiley-Liss, Inc.