Path4Drug

Path4Drug predicts biological pathways affected by drugs by propagating drug–protein interactions to link compounds and target proteins, supporting the prediction of drug-induced adverse effects.

Key Features:

• Integration with Public Repositories: Retrieves drug–target and protein–protein interaction data from ChEMBL, DrugBank, IUPHAR, PharmGKB, TTD, Intact, and MINT to assemble target profiles.
• Pathway Enrichment Analysis: Performs enrichment of target proteins against Reactome to identify pathways potentially influenced by compounds.
• Tissue-Specific Pathway Identification: Applies an optional tissue filter using Human Protein Atlas data to pinpoint tissue-relevant pathway perturbations.
• KNIME Workflow Implementation: Implemented as a KNIME workflow enabling automated retrieval, reconstruction, and processing of drug target data.
• Application to Drug Safety Analysis: Applied to analyze withdrawn drugs and compounds with reported cardio- and hepatotoxic effects as well as cardiac therapy drugs without toxicity records to identify pathways associated with toxic events.

Scientific Applications:

• Systems biology analyses: Linking compounds to pathways to investigate network-level effects of drugs on cellular signaling and regulation.
• Drug toxicity and adverse effect investigation: Identifying pathways and target profiles associated with cardiotoxicity, hepatotoxicity, and other adverse outcomes.
• Post-market and safety profiling: Analyzing withdrawn drugs and drugs with known toxicities to elucidate biochemical pathways underlying safety issues.
• Hypothesis generation for mechanisms of action: Revealing known and novel connections between drugs and biological pathways to support mechanistic interpretations.

Methodology:

Constructs drug target profiles by retrieving protein targets from public databases (ChEMBL, DrugBank, IUPHAR, PharmGKB, TTD, Intact, MINT), propagates drug–protein interactions to link compounds and targets, subjects targets to Reactome pathway enrichment, optionally filters pathways by Human Protein Atlas tissue expression, and implements processing and quality-control via diverse KNIME sub-workflows and filtering steps with automated data retrieval and reconstruction for drugs listed in ChEMBL.