Can LLMs Reliably Annotate Bioassay Metadata to Improve Data Readiness?

TL;DR

Large language models (LLMs) can effectively annotate bioassay metadata, addressing significant gaps in existing databases. This study shows LLMs can enhance data readiness for molecular property prediction.

Key Takeaways

• Evaluate LLMs for annotating bioassay metadata in your datasets.
• Address gaps in assay metadata to enhance data readiness.
• Use LLM insights to identify and correct mislabeled assays.
• Incorporate human review for LLM-generated annotations.