Efficient clustering of identity-by-descent between multiple individuals.

ABSTRACT:

Motivation

Most existing identity-by-descent (IBD) detection methods only consider haplotype pairs; less attention has been paid to considering multiple haplotypes simultaneously, even though IBD is an equivalence relation on haplotypes that partitions a set of haplotypes into IBD clusters. Multiple-haplotype IBD clusters may have advantages over pairwise IBD in some applications, such as IBD mapping. Existing methods for detecting multiple-haplotype IBD clusters are often computationally expensive and unable to handle large samples with thousands of haplotypes.

Results

We present a clustering method, efficient multiple-IBD, which uses pairwise IBD segments to infer multiple-haplotype IBD clusters. It expands clusters from seed haplotypes by adding qualified neighbors and ex

SUBMITTER: Qian Y 

PROVIDER: S-EPMC3967111 | biostudies-literature | 2014 Apr

REPOSITORIES: biostudies-literature