Intelligible Heterogeneous Networks: In-Depth Definition of Lipid Lowering Treatments in Humans

Access to large quantities of diverse data types creates the opportunity to elucidate complex diseases and the heterogeneity of treatments effects. Yet, this opportunity also challenges to render the information ‘big data’ harbours interpretable and directly applicable to patient care. Here, we report intelligible heterogeneous networks defining the complex interactions between molecular effects in liver, plasma, and bile induced by simvastatin and ezetimibe, two of the most prescribed lipid-lowering drugs. The heterogeneous networks incorporate transcriptomics, methylomics, plasma and biliary biochemical parameters from the Stockholm Study, recruiting patients eligible for cholecystectomy and randomized to three active treatments, plus placebo, for four weeks before surgery. Identification of unique heterogeneous modules increases interpretability and leads to the identification of unique putative effects of the different treatments. Proof-of-concept validation in a pre-clinical human model confirms TMBIM6 to be targeted by ezetimibe, alone or in combination, supporting the validity of our network medicine approach at defining biological interactions and discovering new drug targets.