From library to landscape: integrative annotation workflows for compound libraries in drug repurposing

Area of Science:

• Computational chemistry
• Bioinformatics
• Drug discovery

Background:

• Efficient access to chemical and bioactivity data is crucial for drug discovery and repurposing.
• Public repositories like ChEMBL and PubChem contain valuable, yet often disparate, data.

Purpose of the Study:

• To develop automated pipelines for extracting and integrating chemical and bioactivity data from public repositories.
• To create interactive dashboards for multilayered analysis and visualization of drug repurposing data.

Main Methods:

• Developed KNIME- and Python-based annotation pipelines for automated data extraction and integration.
• Utilized ChEMBL and PubChem as data sources, supporting user-provided compound libraries.
• Validated the framework using a harmonized subset of the Specs repurposing collection (>5000 compounds).
• Created interactive dashboards for visualizing chemical properties, bioactivity profiles, and relational data.

Main Results:

• The pipelines successfully automated the collection and harmonization of chemical and bioactivity data.
• The framework enables reproducible workflows for integrating data from heterogeneous sources.
• The Specs subset analysis revealed representative chemical scaffolds of approved and clinical drugs.
• Interactive dashboards facilitate multilayered analyses critical for drug repurposing.

Conclusions:

• The developed framework streamlines data collection and analysis for drug repurposing and discovery.
• The framework is versatile and applicable to broader drug discovery efforts.
• The Specs collection is a valuable resource for identifying drug repurposing candidates.
• Publicly available protocols and open-access dashboards promote wider adoption and collaboration.