Statistical Methods For Haplotype Association Analysis

Genetic studies have shown that the majority of human diseases are the results of genetic and environmental interactions, in which genes play a decisive role. Gene mapping is an important issue in Statistical Genetics, that is, to find the location of the disease locus. With the continuous development of biotechnology, we have received growing SNPs loci, followed by a climbing rate of haplotype categories. Which made us face the high-dimensional, mass, sparse data sets. In this case, modeled either based on a number of points taking the interaction effects between sites into account, or pairs of haplotypes are directly unreasonable. One compromise approach is through the haplotype block methods such as clustering and dimensionality reduction, which would largely reduce the degree of freedom, but also makes the results of a reasonable explanation for the regression parameters. In this paper, we make haplotype association analysis as a starting point, based on complex diseases gene mapping, focusing on the existing haplotypes dimension reduction methods in comprehensive analysis and comparison. It is worth mentioning that, similar to the haplotype such a sparse, massive, high-dimensional data types is also a statistical study of today's hot issues, from life sciences background and the real data actually to explore innovation is regarded as a statistical good strategy.