Assessment of genetic association using haplotypes inferred with uncertainty via markov chain monte carlo

In the last years, haplotypic information has become an important subject in the context of molecular genetic studies. Assuming that some genetic mutations take part in the etiology of some diseases, it could be of great interest to compare sets of genetic variations among different unrelated individuals, inherited in block from their parents, in order to conclude if there is some association between variations and a disease. The main problem is that, in the absence of family data, obtaining haplotypic information is not straightforward: individuals having more than one polymorphic heterozygous locus have uncertain haplotypes. We have developed a Markov Chain Monte Carlo method to estimate simultaneously the sample frequency of each possible haplotype and the association between haplotypes and a disease.