Comparative Assessment of Protein Large Language Models for Enzyme Commission Number Prediction
Crucially, this study demonstrates that LLMs still have to be improved to become the gold standard tool over BLASTp in mainstream enzyme annotation routines. On the other hand, LLMs can provide good predictions for more difficult-to-annotate enzymes, particularly when the identity between the query sequence and the reference database falls below 25%. Our results reinforce the claim that BLASTp and LLM models complement each other and can be more effective when used together.
Authors
- Aalt D. J. van Dijk (ORCID: https://orcid.org/0000-0002-8872-5123)
- Miguel Rocha (ORCID: https://orcid.org/0000-0003-1057-2419)
- Maria Zimmermann‐Kogadeeva (ORCID: https://orcid.org/0000-0001-6021-1246)
- Óscar Dias (ORCID: https://orcid.org/0000-0002-1765-7178)
- João Capela (ORCID: https://orcid.org/0000-0002-1235-2922)
- Dick de Ridder (ORCID: https://orcid.org/0000-0002-4944-4310)
Institutions
- European Molecular Biology Laboratory (DE)
- University of Amsterdam (NL)
- Wageningen University & Research (NL)
- University of Minho (PT)
Publication Details
- Journal
- BMC Bioinformatics
- Published
- 2025-02-27
- DOI
- https://doi.org/10.1186/s12859-025-06081-9
- Citations
- 8
- Primary Topic
- Machine Learning in Bioinformatics
- Type
- article
- Field-Weighted Citation Impact
- 5.64
Funders
- European Molecular Biology Laboratory
- Fundação para a Ciência e a Tecnologia