Large Language Models in Bio-Ontology Research: A Review

Area of Science:

• Life Sciences
• Bioinformatics
• Artificial Intelligence

Background:

• Biomedical ontologies are essential for organizing life science knowledge and enabling integrated analyses.
• Traditional ontology development is time-consuming and requires significant expert input.
• Advancements in artificial intelligence, specifically large language models (LLMs), present opportunities to streamline ontology research.

Purpose of the Study:

• To review recent studies on the application of LLM-assisted techniques in biomedical ontology development.
• To synthesize findings on LLM use in ontology creation, mapping, integration, and semantic search.
• To identify and discuss challenges and future directions in LLM-based bio-ontology research.

Main Methods:

• Literature review of recent studies focusing on LLM applications in biomedical ontologies.
• Synthesis of findings related to LLM-assisted ontology creation, mapping, integration, and semantic search.
• Analysis of challenges including bias, reliability, and ethical considerations.

Main Results:

• LLMs show significant potential to automate and enhance various stages of biomedical ontology development.
• LLM applications span ontology creation, data mapping, integration, and semantic search functionalities.
• Key challenges include managing LLM bias, ensuring reliability, and addressing ethical implications.

Conclusions:

• LLMs are poised to transform the development, maintenance, and utilization of biomedical ontologies.
• Addressing challenges is crucial for the responsible and effective integration of LLMs in bio-ontology.
• Future research should focus on emerging trends to further advance LLM-driven bio-ontology.