A Framework for Classifying Relationships Using Dense SNP Data and Putative Pedigree Information

• Activity Number: 123
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Zhen Zeng*+ and Eleanor Feingold
• Companies: University of Pittsburgh and University of Pittsburgh
• Keywords: genetics ; GWAS ; family relationships ; sequencing ; classification ; SVM

Abstract

When GWAS or sequencing studies are performed on family-based datasets, the genetic marker data can be used to check the structure of putative pedigrees. Even in datasets of putatively unrelated people, both close and distant relationships can often be detected using dense SNP data. A number of methods for finding relationships using genetic data exist, but most are intended for small numbers of uncorrelated genetic markers genotyped on small numbers of pedigrees. Another limitation of existing methods is that many use only a subset of the available information that can be used to classify pedigrees. We propose a set of approaches for classifying relationship types in GWAS datasets or large-scale sequencing datasets. We first propose a method for finding regions of IBD in close relatives using dense SNP data, and then demonstrate how that information can be used in principle to distinguish relationships. We also build an SVM classifier and develop a strategy to take advantage of putative pedigree information to enhance classification accuracy. Finally, we propose a classification pipeline for checking and identifying relationships aimed at datasets containing many small pedigrees.