Empowering drug off-target discovery with metabolic and structural analysis

Area of Science:

• Microbiology
• Computational Biology
• Drug Discovery

Background:

• Identifying intracellular drug targets is a significant challenge in drug discovery.
• Machine learning (ML) analysis of omics data offers a promising avenue but requires refinement to pinpoint specific targets.
• Understanding drug-target interactions is crucial for developing effective therapeutics.

Purpose of the Study:

• To develop and deploy a hierarchical workflow for identifying specific intracellular drug targets.
• To investigate the molecular interactions of the dihydrofolate reductase-targeting antibiotic CD15-3.
• To refine ML-based approaches for drug target discovery, focusing on off-target identification.

Main Methods:

• Utilized a hierarchical workflow combining metabolomics data analysis and growth rescue experiments.
• Applied machine learning, metabolic modeling, and protein structural similarity to prioritize candidate targets.
• Conducted overexpression and in vitro activity assays to validate predicted targets.

Main Results:

• Successfully prioritized candidate drug targets using integrated ML and mechanistic analyses.
• Identified HPPK (folK) as a confirmed off-target for the antibiotic compound CD15-3.
• Demonstrated the effectiveness of the workflow in improving the resolution of drug target identification.

Conclusions:

• The developed hierarchical workflow enhances the precision of drug target discovery by integrating ML with mechanistic insights.
• This approach is effective in identifying off-targets for metabolic inhibitors like CD15-3.
• Combining computational methods with experimental validation is key to advancing drug target identification.