On the inference of ancestries in admixed populations

  • Authors:

  • Sriram Sankararaman;Gad Kimmel;Eran Halperin;Michael I. Jordan

  • Affiliations:

  • Computer Science Division, University of California, Berkeley, CA;Computer Science Division, University of California, Berkeley, CA and International Computer Science Institute, Berkeley, CA;International Computer Science Institute, Berkeley, CA;Computer Science Division, University of California, Berkeley, CA and Department of Statistics, University of California, Berkeley, CA

  • Venue:
  • RECOMB'08 Proceedings of the 12th annual international conference on Research in computational molecular biology
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Inference of ancestral information in recently admixed populations, in which every individual is composed of a mixed ancestry (e.g., African Americans in the US), is a challenging problem. Several previous model-based approaches have used hidden Markov models (HMM) to model the problem, however, the Markov Chain Monte Carlo (MCMC) algorithms underlying these models converge slowly on realistic datasets. While retaining the HMM as a model, we show that a combination of an accurate fast initialization and a local hill-climb in likelihood results in significantly improved estimates of ancestry.We studied this approach in two scenarios--the inference of locus-specific ancestries in a population that is assumed to originate from two unknown ancestral populations, and the inference of allele frequencies in one ancestral population given those in another.