Fusion2AI: functional re-annotation of the bacterial transparent matter proteins
- 1Emory University
Abstract
Determining the biological function of bacterial proteins remains a central challenge in microbiology. We previously developed Fusion, a reference-free scheme that clustered the tens of millions of known, unique, bacterial proteins into about four hundred thousand groups of functionally similar proteins, i.e. functions. For any newly identified proteins, Fusion function assignment could be performed using function-aware sequence alignment to each function's representatives. However, Fusion functions lack human-readable descriptions of their functionality. Here, we leverage large language models (LLMs) to synthesize scattered, partial annotations of functions of individual proteins into coherent English-language descriptions for ~100,000 clusters, i.e. a quarter of all Fusion functions. From these descriptions, we extract Gene Ontology (GO) terms and Enzyme Commission (EC) numbers. We validate these against annotations inferred from UniProtKB/Swiss-Prot. These annotations immediately expand functional coverage for both genome and microbiome -level analyses. For organisms in FusionDB, we annotate 50% more proteins than alignment to Swiss-Prot alone. For other bacteria, the gain is smaller but still reaches 30% for some organisms. For microbiomes, annotation coverage increases nearly ten-fold. Our work thus enables the analysis of microbial "transparent matter," i.e. functions already described in earlier research that have not yet made it into usable annotations. Our annotated Fusion function database is freely available at https://services.bromberglab.org/fusion2ai and as a reference DB for metagenome annotation https://services.bromberglab.org/mifaser.
Related
- GOlien tool: fast and scalable GO term annotation of proteins using a Shannon-entropy k-mer model
- CAZyme Annotation of Five Parasitic Trichomonads Reveals a Conserved Repertoire of Host Glycan Targeting Proteins
- PlantLRR-PRR:A Reproducible Annotation Pipeline Reveals Contrasting Evolution of Developmental and Defense Receptor-Like Proteins in Tomato
- PalmLab: A Comprehensive Computational Platform for Systematic Annotation and Functional Interrogation of Protein Palmitoylation
- A Frozen Plant DNA Language Model for Computational Prioritization and Functional Annotation of Melon (Cucumis melo) Variants
BibTeX
@misc{maloneybertelli2026fusion2ai,
title = {{Fusion2AI: functional re-annotation of the bacterial transparent matter proteins}},
author = {A. R. Maloney-Bertelli and Y. Bromberg},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.28.755017},
url = {https://www.biorxiv.org/content/10.64898/2026.09.28.755017v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - Fusion2AI: functional re-annotation of the bacterial transparent matter proteins
AU - A. R. Maloney-Bertelli
AU - Y. Bromberg
PY - 2026
DA - 2026/09/28
PB - bioRxiv
DO - 10.64898/2026.09.28.755017
UR - https://www.biorxiv.org/content/10.64898/2026.09.28.755017v1
AB - Determining the biological function of bacterial proteins remains a central challenge in microbiology. We previously developed Fusion, a reference-free scheme that clustered the tens of millions of known, unique, bacterial proteins into about four hundred thousand groups of functionally similar proteins, i.e. functions. For any newly identified proteins, Fusion function assignment could be performed using function-aware sequence alignment to each function's representatives. However, Fusion functions lack human-readable descriptions of their functionality. Here, we leverage large language models (LLMs) to synthesize scattered, partial annotations of functions of individual proteins into coherent English-language descriptions for ~100,000 clusters, i.e. a quarter of all Fusion functions. From these descriptions, we extract Gene Ontology (GO) terms and Enzyme Commission (EC) numbers. We validate these against annotations inferred from UniProtKB/Swiss-Prot. These annotations immediately expand functional coverage for both genome and microbiome -level analyses. For organisms in FusionDB, we annotate 50% more proteins than alignment to Swiss-Prot alone. For other bacteria, the gain is smaller but still reaches 30% for some organisms. For microbiomes, annotation coverage increases nearly ten-fold. Our work thus enables the analysis of microbial "transparent matter," i.e. functions already described in earlier research that have not yet made it into usable annotations. Our annotated Fusion function database is freely available at https://services.bromberglab.org/fusion2ai and as a reference DB for metagenome annotation https://services.bromberglab.org/mifaser.
ER -