Novel statistical methods for inference of associations between traits and SNP haplotypes in the presence of uncertain haplotype phase

A single gene can be solely responsible for certain genetic disorders. For example, only people who carry two defective copies of the CFTR gene develop cystic fibrosis. By contrast, complex genetic disorders such as cancer and diabetes likely involve a number of genes that increase susceptibility, and act in conjunction with lifestyle and environmental exposures to increase risk for developing disease. Most success in identifying single causative genes has been achieved by studying co-segregation of a trait with genomic regions in families. However, to tackle complex disorders, researchers have turned from family studies to population studies that investigate associations between a disease and variations in DNA sequences known as single nucleotide polymorphisms (SNPs). Blocks of SNPs, known as haplotypes, offer promise for identifying genes contributing to disease risk. For example, SNP haplotypes were used to help identify a predisposing gene for Crohn's disease. The underlying idea is that similarity among haplotypes of affected individuals will lead to disease associations. Dr. Jinko Graham is developing improved biostatistical methods that account for haplotype uncertainty in analyzing these disease associations. The new techniques will eliminate inaccuracies associated with previous methods and could enable researchers to better evaluate genetic and environmental risks for conditions including diabetes, cancer and cardiovascular disease.