A0186
Title: Accounting for population heterogeneity by modeling interactions with the pliable lasso
Authors: Theophilus Quachie Asenso - University of Oslo (Norway) [presenting]
Manuela Zucknick - University of Oslo (Norway)
Abstract: The pliable lasso penalty is applied to estimate interaction effects and extend the existing linear pliable lasso model to the multi-response problem. In the first part, results from the recent work on the regularized multi-response regression problem are presented, where there exists some structural relation within the responses and between the covariates and a set of modifying variables. To handle this problem, MADMMplasso is proposed, a novel regularized regression method. This method is able to find covariates and their corresponding interactions, with some joint association with multiple related responses. The interaction term is allowed between the covariate and modifying variable to be included in a weak asymmetrical hierarchical manner by first considering whether the corresponding covariate main term is in the model. The results from the simulations and analysis of a pharmacogenomic screen data set show that the proposed method has an advantage in handling correlated responses and interaction effects, both with respect to prediction and variable selection performance. In the second part, results are reported from ongoing work from the implementation of the MADMMplasso in modelling and predicting synergistic effects between two drugs in drug combination experiments, using, for example, the molecular characterization of a cell line with multi-omics data to predict whether two drugs will act synergistically on that particular cell line.
