Human genetic variations and their effect on common complex disease.

摘要

BACKGROUND: The genetic etiology of common complex diseases has been extensively studied during the past few years. Though many causal genetic variants have been identified, they account for only a small percentage of the estimated heritability of complex diseases, such as breast cancer and cigarette smoking. It remains an open question about where the unexplained heritability lies and how to find it. The objective of this dissertation research is to examine three possible sources of such unexplained heritability: 1) the association between copy number variants and breast cancer, 2) the association between gene-gene interactions and cigarette smoking, and 3) the association between functional rare variants and a simulated quantitative trait. METHODS: To detect copy number variants in breast cancer, we examine a breast cancer dataset from the National Cancer Institute and apply a hidden Markov model. To detect gene-gene interactions that are associated with cigarette smoking, we examine a genome-wide dataset from the Study of Addiction: Genetics and Environment and apply a forward U-test. To detect functional rare variants, we examine a dataset from Genetic Analysis Workshop 17 and apply an aggregating U-test. RESULTS: In the breast cancer study, we detect five genomic regions on chromosome 2, 4, 6, 12, and 13. In the cigarette smoking study, we detect two single nucleotide polymorphisms (SNPs) with potential interactions. These two SNPs are located in genes CHRNA5 and NTRK2. In the quantitative trait study, we show that the aggregating U-test has a greater power to detect functional rare variants than a commonly used approach, QuTie. CONCLUSIONS: Our findings from the breast cancer study suggest that structural changes of these genomic regions may contribute to the development of breast cancer. Our findings from the cigarette smoking study indicate that the joint action between genes CHRNA5 and NTRK2 may contribute to the development of cigarette smoking behavior. These proposed methods provide useful tools to detect various types of human genetic variations underlying complex diseases.