De novo inference of stratification and local admixture in sequencing studies.

摘要: Analysis of population structures and genome local ancestry hasbecome increasingly important in disease genetics. With the advance next generation sequencing technologies, complete genetic variants individuals' genomes are quickly generated, providing unprecedented opportunities for learning evolution histories identifying signatures at SNP resolution. The successes those studies critically rely on accurate powerful computational tools that can fully utilize information. Although many algorithms have been developed structure inference admixture mapping, them only work independent SNPs genotype or haplotype format, require a large panel reference individuals. In this paper, we propose novel probabilistic method detecting admixture. takes input data, data data. characterizes dependence via segmentation, such all detected study be utilized inference. further utilizes infinite-state Bayesian Markov model to perform de novo stratification Using simulated datasets from HapMapII 1000Genomes, show our performs superior than several existing algorithms, particularly when limited no individuals available. Our is applicable not human but also other species interests, which little information Software Availability: http://stat.psu.edu/~yuzhang/software/dbm.tar