A multivariate approach to affected-sib-pair analysis using highly dense molecular maps

A multivariate approach to affected-sib-pair analyses was performed to localize disease-susceptibility genes with a minimum number of type I errors (false positives). Using 1,155 independent affected sib pairs extracted from Problem 2A of the GAW10 data set, we were able to localize major genes (MG) 1 and 2. Using 30% of the affected-sib-pair sample (N = 337) we were able to localize MG1. False positives were not detected in either of these samples. © 1997 Wiley-Liss, Inc.