Seminar Abstract

Genetic Association Studies Using Haplotypes

Current practice in association studies of genetic factors underlying complex diseases is to genotype several single nucleotide polymorphisms (SNPs) in a candidate gene. Such data can either be analyzed considering each SNP as a separate predictor, or by considering the effect of haplotypes (the set of SNPs that occur together on the same chromosome). Analyzing haplotypes can have several advantages: they may reduce dimension if there are fewer haplotypes than are combinatorially possible, and because haplotypes reflect the sequence actually transcribed, a haplotype analysis may be more powerful. However, haplotypes are typically unmeasured, and must be inferred statistically. Thus, the problem of analyzing the association between haplotypes and disease outcomes is a type of missing data problem. We consider a variety of new methods for haplotype analysis in association studies, both in case control studies and family-based association studies.

Delivered by Glen A. Satten, National Center for Environmental Health, CDC, Atlanta; April 13, 2004