Measuring gene-disease association using a general pair method

The general-pair-method (GPM) was applied to the analysis of the Problem 1 data set. GPM is a nonparametric, identity-by-state method for associating variation at a chromosomal locus with disease status. It can accommodate information from all pairs of individuals across any degree of genetic relatedness, including genetically unrelated individuals, within a single test statistic. Pairwise contrasts between members of different pedigrees, including "control" pedigrees containing no affected individuals, can therefore be exploited in order to increase the power of the test statistic. Such a strategy makes sense within the context of a candidate gene (direct association) study, where the alleles or genotypes of high disease risk are assumed to be the same ones within all families. Applied to the Problem 1 data set in this way, the GPM successfully identified the effect on disease status of locus 31 on chromosome 1, and of locus 23 on chromosome 5. The observed strength of gene-disease association at the other chromosomal loci was compatible with the null hypothesis.