Under The Framework Of Model Selection Method Based On Bayes Theorem Of Qtl Detection

With the combination of molecular genetics and quantitative genetics, peoplebegan to study QTL mapping. Mapping QTL is the process of estimating thenumber of QTL, their genomic positions, and genetic efects conditional on theobserved phenotype data and marker data. Its researches have greatly influence inbiology, medicine, agriculture and other fields, especially have very greatly help onthe studies of human disease and the economic properties of crops. However, manycomplex traits are afected by multiple QTLs, so the number of the parameters in themodel will be very large, and some traits are difcult to obtain enough samples, inthat case, we can’t use the maximum likelihood method to estimate the parametersin the model.By contrast, the Bayesian method applied in this paper is more easy to dealwith the complex models with multiple QTLs, and it also can be used to estimatethe parameters, which is more than the number of samples, so it is much morewidely used in the QTL analysis of quantitative traits. In the Bayesian analysis,the most important issue is the selection of parameters’ prior. The environmentalfactors will afect the phenotypic trait values, so this paper will use a joint prior toestimate the parameters of population mean and variance. The Bayesian method isimplemented via the Markov chain Monte Carlo algorithm.