MAVEN

MAVEN elucidates drug mechanisms of action by integrating chemical-structure-based target prediction, protein-protein interactions, and transcriptomic perturbation signatures to infer mechanistic sub-networks and perform pathway enrichment using MSigDB and custom gene sets.

Key Features:

• Target prediction and causal reasoning: Performs chemical structure-based target prediction and applies causal reasoning using protein-protein interactions and transcriptomic perturbation signatures to provide a systems-level view of compound MoA.
• Mechanism visualisation: Visualises inferred mechanisms as mechanistic sub-networks that link predicted or known targets to modulated transcription factors through signalling proteins.
• Pathway enrichment analysis: Conducts pathway enrichment on generated networks using MSigDB gene set collections and supports the use of custom gene sets.

Scientific Applications:

• Mechanism of action elucidation: Generates systems-level hypotheses connecting compound targets to downstream transcriptional effects to aid interpretation of pharmacological perturbations.
• Target and pathway identification: Identifies candidate therapeutic targets and affected pathways by integrating chemical-structure-derived predictions with transcriptomic signatures to inform experimental validation.

Methodology:

Integrates chemical structure-based target prediction with causal reasoning derived from protein-protein interaction networks and transcriptomic perturbation signatures to construct mechanistic sub-networks linking targets to transcription factors via signalling proteins, followed by pathway enrichment analysis against MSigDB and optional custom gene sets.