Haplotype recostruction from unphased genotype data at the bovine PRKAG3 gene

Haplotype analysis has became an area of intense research, both for population genetics studies and for molecular dissection of complex phenotypes. Haplotypes provide increased informativeness with respect to single nucleotide polymorphisms and allow to condense information on genomic variation, with a sensible gain of power in association studies. Direct haplotyping via molecular analysis provides more exact information per individual, but is much more expensive and labour-intensive than indirectly inferring haplotypes from genotypes. Several methods have been proposed so far to indirectly reconstruct haplotypes from unphased genotypes. In the present study, we adopted four different approaches (implemented in the computer programs ARLEQUIN, HELIXTREE, HAP and PHASE) to infer phase information from genotypic data on 197 subjects at 14 polymorphic sites of the bovine PRKAG3 gene. In the whole, twelve different haplotypes had been inferred by all the four different methods, although at slightly varying frequencies. Other three haplotypes were inferred by at least two different approaches and some haplotypes were unique to a single method. No dramatic differences among the four selected approaches were observed for the considered genomic target. This is probably due to the linkage-disequilibrium structure of the analysed region and to the moderate amount of missing genotype data