ProtNote: a multimodal method for protein–function annotation
Bioinformatics · 06.05.2025 · 41 (5)
DOI: 10.1093/bioinformatics/btaf170 ↗Kurz gefasst
Fachartikel in der Zeitschrift Bioinformatics, online erschienen am 15.04.2025. Verfasst von Samir Char und 4 weiteren Autor:innen. Beteiligte Einrichtung: Microsoft Cloud & AI, Microsoft, Redmond, WA 98052, United States. Hauptthemen laut MEDLINE: Proteins, Molecular Sequence Annotation, Software, Computational Biology. Die Arbeit wurde bisher 21-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Motivation Understanding the protein sequence–function relationship is essential for advancing protein biology and engineering. However, <1% of known protein sequences have human-verified functions. While deep-learning methods have demonstrated promise for protein–function prediction, current models are limited to predicting only those functions on which they were trained. Results Here, we introduce ProtNote, a multimodal deep-learning model that leverages free-form text to enable both supervised and zero-shot protein–function prediction. ProtNote not only maintains near state-of-the-art performance for annotations in its training set but also generalizes to unseen and novel functions in zero-shot test settings. ProtNote demonstrates superior performance in the prediction of novel Gene Ontology annotations and Enzyme Commission numbers compared to baseline models by capturing nuanced sequence–function relationships that unlock a range of biological use cases inaccessible to prior models. We envision that ProtNote will enhance protein–function discovery by enabling scientists to use free text inputs without restriction to predefined labels—a necessary capability for navigating the dynamic landscape of protein biology. Availability and Implementation The code is available on GitHub: https://github.com/microsoft/protnote; model weights, datasets, and evaluation metrics are provided via Zenodo: https://zenodo.org/records/13897920.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Microsoft Cloud & AI, Microsoft, Redmond, WA 98052, United States.
- Microsoft Cloud & AI, Health & Life Sciences, Microsoft, Redmond, WA 98052, United States.
- Microsoft Research, Microsoft, 1 Memorial Dr, Cambridge, MA 02142, United States.
ORCID-Profile: Kevin K Yang, Ava P Amini
Zitieren
APA
Char, S., Corley, N., Alamdari, S., Yang, K. K., & Amini, A. P. (2025). ProtNote: a multimodal method for protein–function annotation. Bioinformatics, 41(5). https://doi.org/10.1093/bioinformatics/btaf170
Vancouver
Char S, Corley N, Alamdari S, Yang KK, Amini AP. ProtNote: a multimodal method for protein–function annotation. Bioinformatics. 2025;41(5). doi:10.1093/bioinformatics/btaf170
BibTeX
@article{Char2025,
author = {Char, Samir and Corley, Nathaniel and Alamdari, Sarah and Yang, Kevin K and Amini, Ava P},
title = {ProtNote: a multimodal method for protein–function annotation},
journal = {Bioinformatics},
year = {2025},
volume = {41},
number = {5},
doi = {10.1093/bioinformatics/btaf170}
}
EUVIMED