Estimating Local Ancestry in Admixed Populations

摘要: Large-scale genotyping of SNPs has shown a great promise in identifying markers that could be linked to diseases. One the major obstacles involved performing these studies is underlying population substructure produce spurious associations. Population can caused by presence two distinct subpopulations or single pool admixed individuals. In this work, we focus on latter, which significantly harder detect practice. New advances research direction are expected play key role loci different among populations and still associated with disease. We evaluated current methods for inference such cases show they might quite inaccurate even relatively simple scenarios. therefore introduce new method, LAMP (Local Ancestry adMixed Populations), infers ancestry each individual at every single-nucleotide polymorphism (SNP). computes structure overlapping windows contiguous combines results majority vote. Our empirical more accurate efficient than existing inferrring locus-specific ancestries, enabling it handle large-scale datasets. further used estimate admixture individual. experimental evaluation indicates extension yields considerably state-of-the-art as STRUCTURE EIGENSTRAT, frequently correction stratification association studies.