James J Yang1, Jia Li, Anne Buu, L K Williams. 1. Public Health Sciences, Henry Ford Health System, Detroit, Department of Psychiatry, University of Michigan, Ann Arbor and Center for Health Services Research, Henry Ford Health System, Detroit, MI, USA.
Abstract
MOTIVATION: The inference of local ancestry of admixed individuals at every locus provides the basis for admixture mapping. Local ancestry information has been used to identify genetic susceptibility loci. RESULTS: In this study, we developed a statistical method, efficient inference of local ancestry (EILA), which uses fused quantile regression and k-means classifier to infer the local ancestry for admixed individuals. We also conducted a simulation study using HapMap data to evaluate the performance of EILA in comparison with two competing methods, HAPMIX and LAMP. In general, the performance declined as the ancestral distance decreased and the time since admixture increased. EILA performed as well as the other two methods in terms of computational efficiency. In the case of closely related ancestral populations, all the three methods performed poorly. Most importantly, when the ancestral distance was large or moderate, EILA had higher accuracy and lower variation in comparison with the other two methods. AVAILABILITY AND IMPLEMENTATION: EILA is implemented as an R package, which is freely available from the Comprehensive R Archive Network (http://cran.r-project.org/). CONTACT: jyangstat@gmail.com.
MOTIVATION: The inference of local ancestry of admixed individuals at every locus provides the basis for admixture mapping. Local ancestry information has been used to identify genetic susceptibility loci. RESULTS: In this study, we developed a statistical method, efficient inference of local ancestry (EILA), which uses fused quantile regression and k-means classifier to infer the local ancestry for admixed individuals. We also conducted a simulation study using HapMap data to evaluate the performance of EILA in comparison with two competing methods, HAPMIX and LAMP. In general, the performance declined as the ancestral distance decreased and the time since admixture increased. EILA performed as well as the other two methods in terms of computational efficiency. In the case of closely related ancestral populations, all the three methods performed poorly. Most importantly, when the ancestral distance was large or moderate, EILA had higher accuracy and lower variation in comparison with the other two methods. AVAILABILITY AND IMPLEMENTATION: EILA is implemented as an R package, which is freely available from the Comprehensive R Archive Network (http://cran.r-project.org/). CONTACT: jyangstat@gmail.com.
Authors: Alkes L Price; Nick J Patterson; Robert M Plenge; Michael E Weinblatt; Nancy A Shadick; David Reich Journal: Nat Genet Date: 2006-07-23 Impact factor: 38.330
Authors: Alkes L Price; Arti Tandon; Nick Patterson; Kathleen C Barnes; Nicholas Rafaels; Ingo Ruczinski; Terri H Beaty; Rasika Mathias; David Reich; Simon Myers Journal: PLoS Genet Date: 2009-06-19 Impact factor: 5.917