Grant: Operating Grant

Most diseases with serious health impacts such as cancer, cardiovascular disease and psychiatric disorders are called complex due to the large number of genes and environmental factors that combine to cause them. For many complex diseases, comparing the frequencies of differences in the DNA sequence (called genetic variants) between patients and healthy subjects has identified genes contributing to cause them. However, the individual effects of the genes on disease risk are usually small, and researchers will need to jointly analyze other genes and environmental factors with which they interact to accurately predict individual disease risk. It is also recognized that faulty genetic variants may cause impairments associated to a disease. One way to facilitate the detection of disease genes and better understand their action is to measure various traits relevant to a disease in non-affected relatives of patients. A memory impairment in a relative of a schizophrenic patient, or an allergy in a relative of an asthmatic patient is a hint that the relative shares some but not all of the disease-causing variants with the patient. This proposal aims at developing statistical methods to use these hints of disease in two types of gene finding approaches: linkage analysis, which finds regions of the genome transmitted from the parents to their children together with an observable trait, and association analysis, which finds genetic variants overrepresented in people with a trait. The proposed methods model genetic variants in two genes or one gene and one environmental factor simultaneously, and aim the identification of variants making the difference between having the disease or merely an associated impairment. We will also compare different ways of choosing the family members to determine which ones maximize chances of finding such variants. Finding those variants will provide a better understanding of complex diseases to improve preventive and therapeutic interventions.