Traditional linkage analysis and allele-sharing methods locate disease genes using family data. Linkage disequlibrium mapping uses unrelated individuals. Each of these methods has various strengths and weaknesses. I will discuss a new method that could combine some of the strengths of both: searching for shared genomic segments among affected individuals who can be presumed to be distant relatives. This strategy was employed by Houwen et al. (1994) to map a locus associated with benign recurrent intrahepatic cholestasis. They used just three affected individuals from a relatively young isolated population. My talk will focus on the statistical issues that arise in such a study, and in particular on the problem of evaluating false positive probabilities (assessing ``statistical significance") in the face of a great deal of missing information about the population (pedigree) structure.

Back to Workshop Schedule