Thumbnail Image

Publication or External Link





Disease modifiers are genes that when activated can alter the expression of a phenotype associated with a disease. This can be done directly through affecting the expression of another gene that is causing the disease, or indirectly by affecting other factors that contribute to the phenotype’s variability. Identification of disease modifiers is of great interest from both treatment and genetic counseling perspectives. We set here to develop computational approaches to identify and study disease modifiers. We focus on two research avenues for studying disease modifiers: (1) One aimed at identifying and investigating modifiers of cancer, a complex disease influenced by multiple genetic and environmental factors, and (2) the other focuses on the identification of disease modifiers for monogenetic disorders which involve a single disease causing gene.

Towards the first aim of studying cancer modifiers we take four complimentary approaches. (a) First, we developed a computational approach to identify metabolic drivers of cancer that when applied to colorectal cancer, successfully identified FUT9 as a gene that strongly modifies tumors aggressiveness. (b) Second, to study metabolic pathway-level modifications in cancer, we developed an algorithm that summarizes cancer modifications to generate pathway compositions that best capture cancer associated alterations, which, as we show, enhances cancer classification and survival prediction. (c) Third, to identify modifiers of cancer immunotherapy treatment, we developed a new computational approach that robustly predicts the response to immune checkpoint blockage therapy. (d) Fourth, to identify modifiers of cancer radiotherapy treatment we built a robust predictor of rectal cancer patients’ response to chemo-radiation-therapy (CRT), identifying a signature of genes that may serve a potential targets for modifying patients’ response to CRT.

Towards the second aim of studying genetic modifiers of Mendelian diseases, we developed a computational approach for identifying a specific expression pattern associated with genes that are modifying disease severity. We show that we can successfully prioritize genes that are modifying disease severity in cystic fibrosis and spinal muscular atrophy, where we have identified a new modifier and validated it experimentally.

As will become evident from reading my dissertation, my work has naturally focused on developing a variety of computational approaches to analyze research questions that were of interest to me. Obviously, my work has greatly benefited and has been significantly enriched by close collaboration with many experimental labs that have kindly embarked on testing the predictions made, and to whom I am indebted. In sum, we developed methods to identify and study disease modifiers for both cancer and Mendelian diseases. The applications of these methods generates a few promising leads for advancing the treatment for these diseases and improving clinical decision-making.