From The MarthLab
[edit] Computational models of human demographic history based on Population Genetic theory
The genome structure of variations is determined bythe molecular mutation process, random genetic drift modulated by recombination and the effects of long-term demographic history, and the various forms of natural selection. Understanding these primary Biological processes is clearly of interest on its own. They are also of great interest for Medical Genetics because these processes govern allelic association, the non-random assortment between marker and disease allele that makes genetic mapping possible. With quantitative models of allelic association we can make rational decisions for marker spacing in a case-control study. With the knowledge of population specific linkage patterns we can predict how well those markers will work within each population, a question that has been a focus of my research. The two main determinants of allelic association are the fine-scale structure of recombination rates along human chromosomes and demographic history. It is reasonable to assume that recombination rate is governed by molecular processes common to all humans hence the population differences in the strength of allelic association are mainly the consequences of differential demographic histories. Allele frequency data is ideal for studying the effects of long-term demography because the allele frequency spectrum is unaffected by variations in mutation or recombination rates. Using a model fitting approach we determine the model structures and quantify the model parameters that best describe the allele frequency data within each population analyzed.
