Study On Bayesian Quantitative Trait Loci Mapping For Binary Traits And Influencing Factors

A complex binary trait is a character that has a dichotomous expression but with a polygenic genetic background. Mapping quantitative trait loci ( QTLs ) for such trait is difficult because of the discrete nature and the reduced variation in the phenotypic distribution. Bayesian statistics are proved to be a powerful tool for solving complicated genetic problems. In this study, the F2 population of which sample size was 300, information of genetic map and QTLs, the phenotypes of a hypothetical underlying variable ( called the liability ) were simulated by WinQTLCart2.0. The binary trait was linked to liability through a threshold t = 0. The liability was modeled under the classical threshold model and the reversible jump Markov chain Monte Carlo algorithm was used to generate the posterior samples of all unknown parameters for QTLs mapping of binary trait and the posterior means were estimated. Additionally, IM, CIM and MIM were used to map QTLs. The results of bayesian QTL mapping were that the number of QTL was 5, which located in 29.66 cM, 41.94cM, 50.71 cM, 45.75 cM and 84.87 cM respectively and on chromosome 1,2, 3, 4, 4. The additive effects of QTLs were 0.1837, 0.1684, 02407, 0.2050 and 0.4Q22, and their dominant effects were 0.1106, 0.1218, 0.0699, 0.0745 and 0.0575. Comparison with IM, CIM and MIM, the results of bayesian QTL mapping were closer to the true simulated data. Meanwhile, some influencing factors, such as iteration, burn-in and thinning interval, selective genotyping, poisson mean of number of QTL, population size, heritability were analyzed. The results showed that iteration, burn-in and thinning interval had influences on convergence of parameters' chains. The results of selective genotyping were worse than those of genotyping all individuals and their phenotypes because of losing information, which indicated that selective genotyping with an appropriate proportion could be applied for the purposes of both accuracy and economy. Poisson means of number of QTL had influences on the results of Bayesian QTL mapping, which showed that prior distribution affected the posterior means of parameters. With the increasing of population size and heritability, the results of Bayesian QTL mapping were more accurate.