MAVEN: compound mechanism of action analysis and visualisation using transcriptomics and compound structure data in...

Area of Science:

• Drug discovery and development
• Computational biology
• Systems pharmacology

Background:

• Understanding a compound's Mechanism of Action (MoA) is critical for improving drug efficacy and safety.
• Current computational methods for MoA elucidation often require significant coding expertise and are difficult for wet-lab scientists to interpret.
• Existing tools typically focus on predicting direct targets or analyzing downstream pathways, lacking a systems-level, interpretable view.

Purpose of the Study:

• To develop a user-friendly computational tool for predicting and visualizing compound Mechanism of Action (MoA).
• To enable researchers without extensive bioinformatics or cheminformatics knowledge to generate interpretable MoA hypotheses.
• To integrate chemical structure-based target prediction with causal reasoning and transcriptomic data.

Main Methods:

• Development of MAVEN (Mechanism of Action Visualisation and Enrichment), an R/Shiny application with a graphical user interface (GUI).
• Prediction of drug targets based on chemical structure.
• Application of causal reasoning using protein-protein interaction networks and transcriptomic perturbation signatures.
• Generation of a systems-level network view of compound MoA, visualizing links from targets to transcription factors via signaling proteins.
• Inclusion of pathway enrichment analysis using MSigDB gene set collections and support for custom gene sets.

Main Results:

• MAVEN provides a GUI-based approach for predicting drug targets from chemical structures.
• The app generates a systems-level network visualizing the compound's MoA, connecting targets, transcription factors, and signaling proteins.
• MAVEN facilitates pathway enrichment analysis, enabling deeper biological interpretation of compound effects.
• The tool is designed to be accessible to researchers with limited computational expertise, generating interpretable hypotheses.

Conclusions:

• MAVEN is an open-source, user-friendly R/Shiny app that simplifies Mechanism of Action (MoA) elucidation in drug discovery.
• The tool empowers wet-lab scientists to generate testable hypotheses by integrating chemical structure, causal networks, and transcriptomic data.
• MAVEN is available with comprehensive documentation and installation options (Docker, Singularity) for broad accessibility.