Linkage disequilibrium mapping using single nucleotide polymorphisms--which population?

There is considerable interest in the potential of single nucleotide polymorphisms (SNPs) for mapping complex traits which are determined by genes of small individual effect (oligogenes). It is thought likely that many oligogenes are themselves common polymorphisms, perhaps biallelic, for which there is effectively neutral selection reflected in late age of onset. The extent of detectable linkage disequilibrium between SNP x SNP pairs and SNP x oligogene pairs is of considerable interest, particularly in the context of identifying 'favourable' populations. Unfortunately data are sparse and few populations have been extensively sampled. Polymorphisms with the appropriate characteristics that have been studied are blood groups in the Rhesus and MNS systems for which there are extensive data on four pairs of biallelics. These might be regarded as surrogates for SNP-SNP or SNP-oligogene pairs. By developing and applying an approach, previously used for major genes, to evaluate association (rho) in SNP haplotypes, it is evident that, with some exceptions, there is little difference between isolates and large populations. Furthermore it is apparent that there is useful linkage disequilibrium even for the MN-Ss locus pair (0.195 cM apart), in both large populations and isolates. This is somewhat more favourable to linkage disequilibrium mapping than a recent simulation suggests.