Home
 /
Papers
 /
Systems Biology
 /
A protein language model unveils the E. coli pangenome functional landscape regulating host proteostasis
Preprint
Open access
V1
Top 1% · Validated by QED

A protein language model unveils the E. coli pangenome functional landscape regulating host proteostasis

Daniel Martinez-Martinez, Andreea Aprodu, Cassandra Backes +12 more
January 2026
 ·
bioRxiv
 ·
DOI 
10.64898/2026.01.15.699719
94
%
Reproducible
QED Impressions
This paper passed independent validation.
QED's automated review checked the evidence supporting each claim, the statistical methods, and whether the results are reproducible from the data provided.

It ranks in the top 1% of preprints assessed in Systems Biology this quarter.
Run full QED analysis
Read paper
Save

Why this is interesting

Introducing protein language model embeddings to decode pangenome diversity, this work reveals that bacterial chaperones DnaK/ClpX directly tune host ubiquitin-proteasome activity via vitamin B12—overturning the assumption that prokaryotic and eukaryotic proteostasis networks operate independently.

Author perspective

What is the most important insight, discovery, or finding from your paper that you would like other scientists to know?
“

This work represents a conceptual and technical leap in microbiome biology and provides the most comprehensive analysis to date of how strain-level variation within a single bacterial species shapes host molecular outcomes.

We created a bacterial pangenome to host phenotype framework that leverages protein embeddings predicted by the protein large language model ProtT5 through the exploration of 9,558 E. coli genomes allowing us to uncover several novel fundamental principles in microbe-host biology at the strain level.

Finally, our study defines a new conceptual framework of microbe–host co-proteostasis, a phenomenon in which bacterial proteostasis networks directly modulate the host ubiquitin–proteasome system via metabolic communication.

”
Filipe Cabreiro

QED claims analysis

Summary

Protein language model-derived strain embeddings reveal that functional diversity within the E. coli pangenome, rather than phylogenetic relationships, predicts strain-specific host transcriptional responses in C. elegans, with bacterial chaperone DnaK and protease ClpX regulating host ubiquitin-proteasome system activity through coordinated control of vitamin B12 transport and propionate metabolism, establishing a conserved cross-domain co-proteostasis mechanism.

See full analysis
Main Claim 1 –

E. coli Strain-Specific C. elegans Transcriptomes Not Predicted by Phylogeny

Main Claim 2 –

Protein Language Model Embeddings Quantify Pangenome Functional Architecture

Main Claim 3 –

Strain Embedding Landscape Predicts Host Cell Cycle, Metabolism, Proteostasis

Supporting evidence

  • Reporter Assays Confirm Functional Guilds Outperform Phylogeny in Host Regulation
Main Claim 4 –

DnaK and ClpX Tune Host UPS via B12/Propionate

Supporting evidence

  • Bacterial Control of Host Proteostasis Extends Beyond E. coli
  • DnaK-Centered Chaperone–Protease Networks Show Compensatory Proteomic Remodeling
  • Co-Proteostasis Requires Active Bacterial Metabolism to Modulate Host UPS
  • DnaK–ClpX Co-Proteostasis Links TonB and Propionate/BCAA Metabolism
  • TonB Mediates Vitamin B12 Transport in Co-Proteostasis Signaling
  • Vitamin B12 and Propionate Mediate Bacterial Control of Host Proteostasis
  • Bacterial Sleeping Beauty Pathway Drives Propionate Shunt–Dependent UPS Activation
  • Host Propionate Catabolism and Ketone Metabolism Determine Proteostasis Outcome

Abstract

Understanding how bacterial diversity at strain level resolution shapes host physiology is a central challenge in microbiome research. The vast, functionally unknown genetic diversity within a species pangenome makes it difficult to connect genes to function and their impact on host physiology. Here, we explore how the functional landscape of the Escherichia coli pangenome impacts transcriptional responses in Caenorhabditis elegans and show that traditional gene-centric methods fail to provide significant functional associations with the host. Thus, we developed a pangenome framework that leverages the protein language model ProtT5 and generates unique strain embeddings representing the functional potential of each 9,558 E. coli isolate. Stratification of the pangenome into distinct functional guilds aligned with key host processes such as cell division, metabolism and proteostasis. Further, we identify a critical interplay between the extensive network of bacterial chaperones and proteases in regulating host proteostasis. We find that the bacterial chaperone DNAK/HSP70 and protease ClpX fine-tune the host ubiquitin-proteasome system by controlling propionate and vitamin B12 availability. These findings reveal a conserved ‘co-proteostasis’ mechanism as a key phenomenon modulating host-microbe interactions through metabolic communication. Our pangenome-to-phenotype approach offers a powerful strategy to decode bacterial pangenome functional diversity, directly linking microbial genomic variation to host physiological outcomes.

Related papers

How the QED Score works

QED Score is a validated AI-based quality metric that evaluates life science papers on two dimensions: originality and validity. Each manuscript is anonymized, broken down into its claims and supporting evidence, and examined by specialized AI agents for statistical rigor, consistency across figures, and fit with the existing literature. Scores are percentile ranks, so they're comparable across fields. In blinded tests, expert professors sided with QED Score over journal rank in 75% of decisive cases.
Read the methodology

Free access for academic researchers

Create your free QED account to validate your research, strengthen grant proposals, and uncover scientific insights.
We've sent you an access link.
Please check your inbox.

Didn't get your email? Check your spam folder or reach out to info@qedscience.com

Oops! Something went wrong while submitting the form.
Looking for QED for pharma, biotech or life science organizations?