CPIExtract: A software package to collect and harmonize small molecule and protein interactions

Area of Science:

• Biochemistry
• Bioinformatics
• Computational Biology

Background:

• Cellular functions rely on interactions between small molecules and proteins.
• Compound-protein interaction data is fragmented across diverse databases with inconsistent formats, hindering comprehensive analysis.
• Data heterogeneity poses a significant challenge for extracting and integrating experimental binding interaction information.

Purpose of the Study:

• To introduce CPIExtract, a novel tool for interactive extraction, filtering, and harmonization of compound-protein interaction data.
• To overcome data heterogeneity issues and provide a unified, enriched dataset of compound-protein interactions.
• To facilitate downstream applications such as network medicine and machine learning model development.

Main Methods:

• Development of CPIExtract for interactive data extraction from multiple compound-protein interaction databases.
• Implementation of filtering capabilities to refine extracted data based on user-defined criteria.
• Harmonization of data from disparate sources into a unified format.

Main Results:

• CPIExtract successfully extracts and harmonizes experimental compound-protein interaction data.
• The tool demonstrated a >10-fold increase in collected annotations compared to using a single database like DrugBank.
• Output data can be customized via filtering and saved in standard tabular formats.

Conclusions:

• CPIExtract enhances the accessibility and utility of compound-protein interaction data.
• The tool supports advanced analyses, including network medicine for drug repurposing and cross-validation of deep learning models.
• CPIExtract provides a valuable resource for researchers in pharmacology and computational biology.