A0440
Title: Bayesian nonparametric mixtures of categorical directed graphs for heterogeneous causal inference
Authors: Federico Castelletti - Università Cattolica del Sacro Cuore (Milan) (Italy) [presenting]
Laura Ferrini - Universita Cattolica del Sacro Cuore (Italy)
Abstract: Quantifying the causal effects of exposures on outcomes, such as a treatment and a disease, respectively, is a crucial issue in medical science for the administration of effective therapies. Importantly, any related causal analysis should account for all those variables, e.g. clinical features, that can act as risk factors involved in the occurrence of a disease. In addition, the selection of targeted strategies for therapy administration requires quantifying such treatment effects at a personalized level rather than at a population level. These issues are addressed by proposing a methodology based on categorical directed acyclic graphs (DAGs), which provide an effective tool for inferring causal relationships and causal effects between variables. In addition, population heterogeneity is accounted for by considering a Dirichlet process mixture of categorical DAGs, which clusters individuals into homogeneous groups characterized by common causal structures, dependence parameters and causal effects. Computational strategies are developed for Bayesian posterior inference, from which a battery of causal effects at the subject-specific level is recovered. The methodology is evaluated through simulations and applied to a dataset of breast cancer patients to investigate side effects associated with the occurrence of cardiotoxicity and possibly implied by the administration of anticancer therapies.
