NetMe 2.0: a web-based platform for extracting and modeling knowledge from biomedical literature as a labeled graph

Area of Science:

• Bioinformatics
• Computational Biology
• Biomedical Informatics

Background:

• The exponential growth of biomedical literature presents a significant challenge for researchers seeking to stay current with scientific advancements.
• Network analysis is a vital computational tool in life sciences, but constructing accurate and comprehensive biomedical networks from literature remains difficult.

Purpose of the Study:

• To introduce NetMe 2.0, a web-based platform designed to automate the construction of biomedical knowledge graphs (BKG) from diverse text sources.
• To enhance the utility of BKGs through an innovative Retrieval Augmented Generation (Graph-RAG) module for generating explanatory sentences.

Main Methods:

• NetMe 2.0 automatically extracts biomedical entities and their relationships from user-provided texts (PubMed Central articles, free text, PDFs).
• Extracted information is modeled into a BioMedical Knowledge Graph (BKG).
• A Graph-RAG module processes the BKG to generate coherent sentences explaining the modeled relationships.

Main Results:

• NetMe 2.0 successfully infers comprehensive and reliable biological networks.
• The platform demonstrates significant Precision-Recall metrics when compared to existing state-of-the-art methods.
• The Graph-RAG module effectively distills well-formed sentences explaining the BKG content.

Conclusions:

• NetMe 2.0 offers a robust solution for overcoming the challenges of building biomedical networks from extensive literature.
• The platform facilitates a deeper understanding of complex biological relationships through automated knowledge graph construction and explanation generation.
• NetMe 2.0 is available at https://netme.click/.