In the course of the study about animals and plants, binary trait is one of usualtraits, e.g, a individual is afected or not. Binary trait locus mapping, which mapsloci in the genome that afect a binary trait, is of important scientifc and practicalvalue. Three improved methods of binary trait locus mapping are introduced inthis paper. Selective genotyping is a method with extreme sib-pair on the samplechoosing. And the cost of gene mapping can be reduced. Logistic regression mixturemodel is a method of improved phenotype model, which is with random efect andmakes the mapping have more practical meaning. Model selection is improved onthe using of makers, and it can be greatly improved the efciency of gene mapping.On the feld of multivalue trait locus mapping, multiple-interval mapping(MIM) andbayesian mapping for ordinal traits are introduced. MIM is higher power and facilityto analysis epistasis. The advantage of bayesian mapping is that we can stepwisemodel select and estimate the parameters at the same time. Through the studyingand analysing the mapping methods above, we proposed an strategy to map thedisordinal and multivalue traits, then it was simulated and analysed. |