Personal tools

GOLD - Genome Wide Association


Jump to: navigation, search
Title Fast and exact multiple testing correction for millions of correlated genetic markers in large sample size studies
Advisor Mohammed S. Al-Rawi
Level First
Target genome-wide association studies GWAS, other multiple correction models
Area science and engineering
See also

In the last decade, the topic of genome–wide association studies (GWAS) has received worldwide attention and considerable funding, e.g. by remarkable agencies such as NIH (national institute of health, USA), MRC (medical research center, UK), have been exclusively dedicated for researchers working on it. GWAS can test millions of single nucleotide polymorphism (SNP) markers for association with a trait. Since SNPs in linkage disequilibrium (LD) are correlated to some degree, a multiple testing correction method is needed. Using Bonferroni or Sidak correction is typically too stringent multiple testing in GWAS. Permutation testing that is implemented as multiple testing correction method has been considered the gold-standard in determining the adjusted p-values in GWAS. For a study consisted of N individuals, M markers, and K permutations (which should be high enough to ensure high accuracy), permutation testing requires heavy computations of the order O(NMK), which is very slow and high memory demanding. Our work aims at reducing the computational complexity of permutation testing in GWAS problems. The greatest issue in our method is that it is independent of the sample size (N), thus, would be best suited in large sample size studies.!forum/project_gold