Overview
A Statistical Framework for Quantitative Trait Mapping
QTL Archive
QTL Mapping Overview
Simple genetic traits and diseases represent only a small portion of the total genetic contribution to human health problems. Most traits with great impact on human health are complex, involving many genes that can interact with one another and with environmental factors, making the prediction of disease state for a given genotype a difficult task. We plan to use crosses between of inbred strains of mouse as a model system to study the complex genetics underlying common diseases including diabetes, obesity, heart disease, hypertension and osteoporosis.
The common approach (e.g., MapMakerQTL) of testing one locus at a time is not able to detect genes that are important by virtue of epistatic interactions. We have taken a first step in this direction by developing methods for simultaneous genome scans of all possible pairwise interactions. In almost every instance we have looked at, there is significant evidence for gene x gene interaction. In many cases the interaction terms dominate the explanatory power of the model and in some cases genes of major effect are not detected by single gene approaches. When gene interactions are driving a phenotype, one can propose experimentally verifiable hypotheses regarding the role of these genes in pathways. We have already begun construction of combinatorial congenic strains to prove the roles of interacting loci uncovered by our analysis of atherosclerosis susceptibility in one cross and bone density parameters in another.
A second major direction that our QTL work has taken is to look at the concordance of QTL findings across species. In the case of hypertension, we are finding a remarkable degree of concordance of QTL in comparisons of mouse with rat and human.
QTL mapping is a phenotype driven approach to gene function. As such it permits the discovery of new genes and can be contrasted with gene-driven approaches such knock-out and knock-in mice which allow for the study of known genes. QTL reflect natural genetic variations as they exist in the mouse strains under study. We are limited to detecting those genes that vary among the available strains. However the natural variations among mouse strains are vast and largely untapped. Another phenotype driven approach, ENU mutagenesis also holds great promise for revealing function of many genes. However we are concerned that ENU mutants are not likely to reflect natural variation. We believe that QTL mapping in model organisms will continue to play an important role in elucidating the relationship between genotype and phenotype in complex systems and that it will be a valuable tool for understanding the role of complex genetics in human health.
