ThermoMutDB 2.0: semiautomated literature curation and structural enrichment of thermostability data for protein mutations.

ThermoMutDB 2.0 screened 5,934 candidate publications (1976–2026) and, after expert review, incorporated 102 articles, adding 1,763 new mutation–thermostability entries.

• Why it matters: Protein thermodynamic stability measurements are crucial, but extracting mutation effects from the literature is labor-intensive, limiting timely, structured access to experimentally grounded data.
• What they did: The framework combines a large language model with rule-based entity recognition to triage and prioritize papers for expert validation, then enriches curated mutations with mutant structural models for 3D context.
• The result: The integrated human-in-the-loop workflow enables sustainable continuous maintenance and links experimentally measured protein stability changes to structural context for interactive exploration.

AI summary of the abstract · may contain errors · not medical advice · read the paper.