Haplotyping for linkage disequilibrium mapping

Abstract

The underlying idea of Linkage Disequilibrium (LD) mapping is that genotyped loci can act as a proxy for ungenotyped loci within haplotypes. Consequently LD mapping can exploit recombinations that occurred in ancestors of the sampled population. LD mapping may use haplotypes directly, or indirectly by using identity by descent (IBD) matrices. Haplotypes - sets of linked alleles segregating together - are essential for LD mapping. It is possible to determine haplotypes using molecular techniques to sequence small sections (a few kilobases) of chromosomes (Tost et al., 2002, Mitra et al., 2003) and also for complete chromosomes using a process called conversion (Douglas et al., 2001). Conversion transforms diploid cells to haploid cells by constructing somatic cell hybrids, some of which end up with single copies of some chromosomes. These can be identified using a small number of markers and genotyped to produce accurate haplotypes for whole genomes. However, these methods are expensive and unlikely to find wide use. Large quantities of genotypic information, in the form of dense bi-allelic genotypes (Single Nucleotide Polymorphisms, SNP), are becoming available to researchers. These data come as unordered (phase-unknown) genotypes so that the parental source of each allele is unknown. The process of haplotyping is to reconstruct gametes from diploid genotypes and can also be seen as ordering genotypes.