Conference is now cancelled
A functional regression model for determining drug-response relationship in cancer genetics
Evanthia Koukouli (Lancaster University)
Cancer treatments can be highly toxic, and frequently only a subset of the patient population will benefit from a given treatment. There is evidence that tumour genetic mechanisms influence cancer drug sensitivity. We suspect that specific genetic factors could be used as a decision aid for treatment selection or dosage tuning. Experiments using human cancer cell lines treated with known anti-cancer compounds can enlighten researchers in this regard. Using dose-response and gene expression data from the Genomics of Drug Sensitivity in Cancer (GDSC) project, we build a dose-varying regression model. We investigate the effectiveness of five BRAF targeted compounds applied to different cancer cell lines under different dosage levels. We assume that out of tens of thousands of genes only a small proportion are actually associated with cancer cell survival in a dosage-dependent manner. We adjust for the gene expression profiles by including them as dose-invariant covariates into the regression model, and assume that their effect varies smoothly over the dosage levels. A two stage variable selection algorithm (variable screening followed by penalised regression) is used to identify genetic factors that are associated with drug response over the varying dosages. We evaluate the effectiveness of our method using simulation studies focusing on the choice of tuning parameters and cross-validation for predictive accuracy assessment. We give examples of the relationships between the selected genes and drug response and we perform a pathway analysis to investigate the biological importance of our findings.