UC San Diego

Genetic variants, or changes in DNA sequence, are known to contribute to both complex and Mendelian diseases. The identification of individual and collections of variants, both common and rare, associated with diseases can help elucidate pathogenic mechanisms contributing to those diseases since it is known that genetic variants can impact gene function and drive pathophysiology. Unfortunately, there is no consensus on the best strategies for identifying genetic associations and effects. In fact, many methods simply involve testing each variant in the genome for association with a trait directly and ignore the fact that most molecular and physiological systems are quite complex and involve a number of interacting parts. In this light, the effect of any one variant may be masked by, or interact with, other variants and phenomena (such as environmental factors). This is a likely reason why many attempts to identify genetic variants associated with most diseases have not been able to explain the majority of the heritable component of those diseases. It is, therefore, important to consider genetic association analysis methods that are sensitive to the fact that genetic variants may exhibit effects that are ``context dependent'' in that their effects depend on the existence of other variants or environmental factors.

Quantifying the extent to which genetic variants interact with other factors remains a challenge in genetic studies. This is the case despite the fact that there have been numerous historical studies exposing the existence of context dependent genetic effects in very broad settings that should motivate greater concern for context dependency in modern genetic association studies. For example, many model organism studies, highly contrived in vitro studies, studies of tumor responsiveness to targeted therapies, and general clinical studies of monogenic diseases have all suggested that the phenotypic impact of certain genetic factors is dependent on other factors. We believe that ignoring the genetic and overall context within which a genetic variant is operating can negatively impact understanding disease pathogenesis and human biology.

In the following, we explore two broad settings in which genetic background and context can have an effect on the interpretation of the impact of genetic variation on a clinically meaningful phenotype. The first setting involves associating genetic variation exhibited by the pathogen Methicillin-Resistant Staphylococcus Aureus (MRSA) and the clinical outcomes of patients harboring an infection induced by that pathogen. Essentially, the current manner in which MRSA genetic variants are identified requires the choice of a reference strain genome whose genetic background relative to the strains of interest could influence the characterization, association and interpretation of the impact of those variants. The second setting considers the identification of genetic factors that either collectively influence Alzheimer's Disease (AD) in a manner that is dependent on the genetic background of the individuals studied or that work through mechanisms that can lead to their association with AD only if that mechanism is explicitly modeled. We ultimately believe that the approaches and findings in our work should motivate further research and a sensitivity to the numerous contexts in which genetic variants may impact a phenotype.