Prototyping CRISP: A Causal Relation and Inference Search Platform applied to Colorectal Cancer Data

Abstract

We introduce CRISP, a Causal Research and Inference Search Platform. It is designed to assist biological and medical research by applying a variety of causal discovery methods to heterogeneous and high-dimensional observational data. CRISP aims to identify a small set of input variables which are most likely to have a causal effect on a target variable. The output of CRISP, thus, highlights the most promising candidates for further targeted research. We illustrate the utility of CRISP with a case study in oncology, using a multi-omic colorectal cancer data set to identify causal drivers differentiating two subtypes of colorectal cancer.