Molecular network analysis of hormonal contraceptives side effects via database integration

Abstract

Hormonal contraceptives (HCs) have been shown to be safe and effective when used correctly and consistently, however, as other classes of drugs, they are also associated with adverse health outcomes. In this study, we aim to explain the occurrence of common and unexpected HCs side effects (SEs) integrating drug-target, drug-SE and protein-protein interaction (PPI) public databases. We created a tripartite network that includes three types of vertices: SEs, drugs, and targets. The three layers are linked by means of the inter-layer associations drug-target and drug-SE, whereas only the target layer is characterized also by intra-layer links (PPIs). We exploited the drug-mediated association SE-target to identify the side effect modules defined as a network connected component composed of target proteins plus the proteins needed to connect them. We found that module proteins are associated with diseases/phenotypes and/or KEGG pathways related to the SEs. In particular, in many cases, targets are not enriched in SE features, whereas investigating their neighborhood (here defined as the proteins that allow the targets' connection) we found SE-related pathways. These results show that HCs action can perturb the targets’ neighborhood inducing unwanted reaction and that the proposed approach can help to understand how, and through which molecular mechanisms, side effects can occur. The approach is general in its nature: it can be applied to other drugs categories providing a support in identifying a subject-specific therapy that takes into account comorbidities and lifestyle to reduce or avoid the most undesired side effects.

Keywords

• Hormonal contraceptives
• Multipartite graph
• Network medicine
• Side effect module
• Precision medicine