From Knowledge Generation to Knowledge Verification: Examining the BioMedical Generative Capabilities of ChatGPT

2025-02-20

AI Summary

This work addresses the challenge of evaluating factual accuracy in knowledge generation by large language models (LLMs) for biomedical associations—particularly disease–drug, disease–gene, and disease–symptom relationships. We propose a generative–verificative two-stage framework that decouples knowledge generation from rigorous validation. Our method introduces disease-centric prompt engineering and integrates biomedical ontologies—including DOID, ChEBI, and GO—to enable terminology standardization, ontology alignment, and semantic-level verification. Experiments show high accuracy in identifying disease (98%), drug (92%), and gene (88%) entities, with literature support rates of 89%–91%. In contrast, symptom identification accuracy is markedly lower (49%–61%), revealing for the first time a critical limitation of LLMs in symptom modeling. This study establishes an interpretable, reproducible, and quantitative paradigm for trustworthy biomedical knowledge generation and evaluation.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The generative capabilities of LLM models present opportunities in accelerating tasks and concerns with the authenticity of the knowledge it produces. To address the concerns, we present a computational approach that systematically evaluates the factual accuracy of biomedical knowledge that an LLM model has been prompted to generate. Our approach encompasses two processes: the generation of disease-centric associations and the verification of them using the semantic knowledge of the biomedical ontologies. Using ChatGPT as the select LLM model, we designed a set of prompt-engineering processes to generate linkages between diseases, drugs, symptoms, and genes to establish grounds for assessments. Experimental results demonstrate high accuracy in identifying disease terms (88%-97%), drug names (90%-91%), and genetic information (88%-98%). The symptom term identification accuracy was notably lower (49%-61%), as verified against the DOID, ChEBI, SYMPTOM, and GO ontologies accordingly. The verification of associations reveals literature coverage rates of (89%-91%) among disease-drug and disease-gene associations. The low identification accuracy for symptom terms also contributed to the verification of symptom-related associations (49%-62%).

Problem

• Evaluate factual accuracy of biomedical knowledge
• Verify disease-drug-gene-symptom associations
• Assess ChatGPT's generative capabilities in biomedicine

Innovation

• LLM model evaluation
• biomedical ontology verification
• prompt-engineering processes