Resolving Complex Traits via Coancestry in Pedigrees and Populations

• Activity Number: 214
• Type: Invited
• Date/Time: Monday, July 30, 2012 : 2:00 PM to 3:50 PM
• Sponsor: International Indian Statistical Association
• Author(s): Elizabeth Alison Thompson*+
• Companies: University of Washington
• Keywords: Genetics ; Coancestry ; Hidden Markov models ; Latent variables ; Pedigree structures

Abstract

Gene identity by descent (ibd) underlies all patterns of phenotypic similarity among relatives, whether these relationships are close and known, as in pedigrees, or more remote and unknown, as in a population. At any locus, the pattern of ibd among individuals observed for a trait of interest, defines the ibd-graph. All inference relating to the trait contributions at this locus depend only on the ibd graph, and not additionally on the pedigree or relationship structure among the individuals. On a base-pair scale, ibd-graphs are slowly varying along a chromosome.

In population data, segments of ibd can be accurately inferred from modern dense SNP marker data, while within a pedigree the small components of the ibd graph can often also be quite precisely inferred. Moreover, the within-pedigree data permit accurate haplotyping of observed individuals, which increases the accuracy of between-pedigree ibd inference. We demonstrate how, by combining the ibd graphs within and between the pedigrees in a study, the power and resolution of genetic linkage studies can be enhanced.