GePI: large-scale text mining, customized retrieval and flexible filtering of gene/protein interactions

Area of Science:

• Bioinformatics
• Computational Biology
• Biomedical Informatics

Background:

• Scientific literature contains vast amounts of information on molecular interactions.
• Extracting and organizing this information manually is time-consuming and challenging.
• There is a need for automated tools to mine and access gene and protein interaction data.

Purpose of the Study:

• To develop a novel web server, GePI, for large-scale text mining of molecular interactions.
• To provide researchers with efficient tools for retrieving and contextualizing gene and protein interaction data.
• To offer an up-to-date and user-friendly platform for monitoring molecular interaction information.

Main Methods:

• Leveraging natural language processing (NLP) techniques to identify genes, related entities, and biomolecular events.
• Implementing powerful search options with full-text filters for contextualized queries.
• Developing a knowledge graph updated multiple times a week for current data.
• Designing a results page with interaction statistics, visualizations, and downloadable Excel tables.

Main Results:

• GePI successfully identifies genes, entities, and their interactions from biomedical literature.
• The server supports rapid retrieval of interactions with flexible search and filtering capabilities.
• Results include interaction pairs, molecular entities, factual certainty, and supporting text snippets.
• The knowledge graph is kept current through frequent updates.

Conclusions:

• GePI offers a free, easy-to-use, and up-to-date solution for monitoring gene and protein interactions.
• The tool enhances biomedical research by providing efficient access to extracted interaction data.
• Flexible query formulation and filtering options make GePI a valuable resource for scientists.