Characterizing Genome Structure of Admixed Populations for Disease Association Studies
Neil A. Hattangadi
HST MD 2005
David Altshuler, MD, PhD
Whitehead Institute Center for Genome Research
Massachusetts General Hospital, Molecular Biology Department
It is believed that common genetic variants play a significant role in susceptibility to common diseases, such as hypertension, cancer, and autoimmune disorders.1 The difficulty of finding these common disease-associated polymorphisms arises because the disease phenotypes are complex, with a strong environmental component, and the genetic contribution is multiallelic with varying, incomplete penetrance.2
The approach we are using to identify disease-associated polymorphisms is to analyze historically divergent populations with varying disease rates. A number of diseases show large differences in prevalence between European and African populations; for example, multiple sclerosis is much more common in Europeans and prostate cancer in Africans. We can search for markers that are associated with these diseases by studying the genomes of African-American individuals, which represent the admixture of European and African genomes. We compute the probability that a given chromosomal region of an individual African-American genome descends from European or African ancestry. Regions of the African-American genome which are enriched for European ancestry in patients with multiple sclerosis, but not in those without the disease, may be associated with MS; the converse is true for prostate cancer.
In this project, a quantitative genotyping methodology has been developed which permits rapid identification of ancestry-informative markers. We have also developed a Hidden Markov Model to measure the size of "ancestry blocks" – regions of the African-American genome of continuous ancestry due to linkage disequilibrium. We have found the ancestry blocks to be in excess of 15 Mb, suggesting far fewer markers are needed for admixture-based disease studies than conventional genetic association studies. We also use the HMM to compute the expected ancestral origin at any point on the admixed chromosome. The model is being applied to two large case-control populations for multiple sclerosis and prostate cancer to identify potential disease-associated loci.
1Lander, E.S. (1996) The new genomics: global views of biology. Science 274, 536-539.
2Altshuler, D., et al. (2000) Guilt by association. Nature Genetics 26, 135-137.
3Stephens, J.C., et al. (1994) Mapping by admixture linkage disequilibrium in human populations: limits and guidelines. Am. J. Hum. Genet. 55, 809-824.
Previous
Next
AI
