Enhancing pharmacogenomic data accessibility and drug safety with large language models: a case study with Llama3.1
2024

Using Large Language Models to Improve Drug Safety and Pharmacogenomic Data Access

Sample size: 178 publication 10 minutes Evidence: high

Author Information

Author(s): Li Dan, Wu Leihong, Lin Ying-Chi, Huang Ho-Yin, Cotton Ebony, Liu Qi, Chen Ru, Huang Ruihao, Zhang Yifan, Xu Joshua

Primary Institution: National Center for Toxicological Research, U.S. Food and Drug Administration

Hypothesis

Can Large Language Models (LLMs) automate and enhance the accuracy of pharmacogenomic data extraction?

Conclusion

The study found that Llama3.1-70B achieved high accuracy in extracting pharmacogenomic information, demonstrating its potential to streamline data extraction processes.

Supporting Evidence

  • Llama3.1-70B achieved 91.4% accuracy in identifying drug-biomarker pairs from structured texts.
  • The model demonstrated 82% accuracy in extracting information from mixed texts.
  • The study highlighted the importance of integrating diverse data sources for comprehensive pharmacogenomic insights.
  • The model effectively identified ethnicity differences in drug response across various populations.

Takeaway

This study shows that a computer program can help doctors find important information about how people's genes affect their responses to medicine, making it easier to give the right treatment.

Methodology

The study used the Llama3.1-70B model to extract pharmacogenomic data from the FDA PGx Biomarker table and relevant scientific abstracts.

Potential Biases

Potential bias in the representation of minority groups in pharmacogenomic research.

Limitations

The model struggled with less common biomarker names and mixed texts, which may affect extraction accuracy.

Participant Demographics

The study included data related to various ethnic groups, including Caucasians, African Americans, and Asians.

Statistical Information

P-Value

0.0001

Confidence Interval

Not specified

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.3389/ebm.2024.10393

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication