Before performing the linkage analyses, we began by making sure that the data had the properties we expected. We believe that this is an important but often neglected step for both real and simulated data. While attempting to identify the population origin of the 202 founders of the 8 pedigrees, we discovered to our surprise that 4,239 SNPs had been altered, although these alterations did not prevent the unambiguous identity and population origin of the founders. We were even more surprised to discover that the founders were not a random sample from the unrelated individuals data set. Instead, all but three of the founders were drawn from the two African samples. Moreover, three of the eight pedigrees were established by approximately equal numbers of founders from the Yoruba and the Luhya population samples. The remaining five pedigrees were primarily founded by one ethnic group or the other. Under ordinary circumstances when using genetic markers for a linkage analysis, it is more cumbersome to analyze families composed of individuals drawn from various ethnic groups because differences in allele frequencies need to
