Models For Multi-locus Linkage Analysis In Cross Population Of Tetraploids

Polyploidy is pervasive in plants and some estimates that more than50%of angiosperms are polyploidy, or experienced genome doubling in evolutionary history. Due to its special features, polyploidy not only display significant value in research on plant genetic evolution, but also in agricultural breeding practices. Linkage analysis is a statistical method for estimating the recombination fraction and genetic distance between markers. Genetic map, which is based on linkage analysis, is one of the most important tools in the field of genetics study. However, because of some special meiotic behavior---double reduction and preferential pairing, traditional method for estimating linkage cannot be deployed in polyploidy population.In this thesis, two models were developed for linkage analysis in polyploidy species. These two models based on three points analysis and incorporated the important meiotic feature of polyploidy. By implementing the EM algorithm within the maximum likelihood framework, the models provides a procedure for simultaneous estimation of linkage and parameters measuring the meiotic behavior in polyploidy.A series simulation studies and real data analysis has demonstrated the new models for polyploidy species can be well applied within fully informative markers and partially informative markers. Compare to the pervious method baseds on two points analysis, these two three points analysis method could provide more accurate estimation of recombination fraction.The models provide robust statistical tools for linkage analysis in polyploidy population. The special meiotic behavior of polyploidy are modeled and measured in the models, therefore, the result from new models will help shed light on genetic diversity and origin of a polyploid population.