bio.rodeo
ModelsOrganizationsProvidersLeaderboardAboutSign in
bio.rodeo

The authoritative source for evaluating biological foundation models. No hype, just honest analysis.

Categories
  • DNA & Gene
  • RNA
  • Protein
  • Small molecule
  • Single-cell
  • Spatial omics
  • Pathology
  • Imaging
  • Metabolomics
  • Biosignals
  • Language model
bio.rodeoModelsOrganizationsProvidersLeaderboardAboutFAQSubmit a modelContact
© 2026 Pulsatance. All rights reserved. ~
Built by Pulsatance
models / dna-gene / pide
DNA & GeneProtein
Tsinghua UniversityReleased March 2025

PIDE

Prophage island detection in bacterial genomes and metagenome-assembled genomes, pairing a fine-tuned ESM-2 gene classifier with density clustering.

The short version

  • —Resolves prophage boundaries at nucleotide resolution, checked against induced VLP reads
  • —Classifies every predicted ORF by protein language, so diverged prophages still register
  • —Density clustering stitches phage genes across the bacterial cargo genes islands carry
  • —Also sorts phage contigs from bacterial ones in metagenome assemblies down to 0.5 kb
56Openness
23GitHub stars

Where to run it

No providers recorded yet. Browse all providers

PIDE — Prophage Island Detection using ESM-2 — finds integrated phage genomes inside bacterial chromosomes and metagenome-assembled genomes. It was developed at the School of Medicine at Tsinghua University posted as a preprint in early 2025 and published in Genome Biology in August 2025.

Prophages are a large and consequential fraction of bacterial genomes: they carry toxin and antibiotic resistance genes, mediate horizontal transfer, and reshape host populations when they excise. Finding them computationally is hard for two related reasons. Phage proteins evolve fast, so similarity searches miss diverged elements entirely; and even when a region is flagged, existing tools tend to be imprecise about where it starts and stops, because prophages interleave with bacterial cargo genes rather than sitting as clean blocks.

PIDE addresses both problems by moving the decision to the protein level and the boundary call to a separate step. Every open reading frame is embedded with a fine-tuned ESM-2 and classified as phage or bacterial, which recognizes phage proteins that have no close match in any reference database; adjacent phage calls are then merged into islands by gene-density clustering, which lets a prophage keep its interspersed bacterial genes without losing its edges. The result sits alongside broader metagenomic virus finders such as DeepVirus, but is aimed specifically at the boundary problem.

#Key Features

  • Protein language classification of every ORF: Prodigal-called ORFs are scored individually by a protein language model rather than matched against a phage reference database, so novel and diverged prophage proteins are still recognized.
  • Density clustering with score-based trimming: Adjacent phage ORFs within a configurable intergenic distance are merged into one island, and ORFs are iteratively trimmed from both ends until the island's mean phage probability clears the threshold.
  • Nucleotide-level boundary accuracy: On viral-like-particle sequencing from induced cultures, PIDE reached a mean base precision of 0.91, ahead of geNomad at 0.81, PHASTER at 0.70, and VirSorter2 at 0.35, while holding base recall at 0.91.
  • Metagenomic contig mode: The same ORF classifier separates phage contigs from bacterial ones in assemblies, with thresholds tuned per contig length from 0.5 kb up to 20 kb.
  • Sensitivity to elements other tools miss: Across 38 gut bacterial isolates, PIDE called 365 islands to geNomad's 142, PHASTER's 174, and VirSorter2's 101, and 184 of those were unique to PIDE.

#Technical Details

The classifier uses ESM-2 at 650 million parameters and 33 transformer encoder layers. Per-residue embeddings are mean-pooled into a 1,280-dimensional vector and passed to a five-layer perceptron with ReLU activations and a softmax output; ORFs scoring above 0.5 are called phage. Only the last four encoder layers are fine-tuned while the rest stay frozen, and the perceptron is trained from scratch, with learning rate 5e-6, batch size 4, and weight decay 1e-5, converging in two epochs. Training data came from UniRef50 (May 2024): 263,843 phage proteins as positives, and an equal number sampled from 40.5 million bacterial and archaeal proteins after removing anything above 95% identity to the phage set with cd-hit-2d.

On the held-out split the classifier reached accuracy 0.90, F1 0.90, AUC 0.96, precision 0.86, recall 0.95, and average precision 0.95. An ablation without fine-tuning ESM-2 lost ground chiefly in recall, at 0.93, confirming that adapting the language model rather than probing it frozen is what buys the sensitivity. Island assembly defaults to a 3,000 bp clustering distance, an island score threshold of 0.7, and a minimum of five phage ORFs.

#Applications

Microbiome researchers annotating bacterial isolates or assembling genomes from stool, soil, or ocean metagenomes can use PIDE to inventory the mobile fraction of those genomes and to catalog what it carries. The authors ran it across 4,744 human gut species representatives, detecting 24,467 prophage islands in 4,198 genomes — 88.5% of them — with a median island of 7,738 bp and an average of 2.8% of host genome content. Of those lysogens, 96.2% carried cargo genes, and 34.2% of the bacteria harbored islands carrying 120 distinct antibiotic resistance gene types across 18 categories, a direct readout of the mobile resistance reservoir in the gut.

#Impact

PIDE's contribution is precision at the edges. Prior tools either overcall regions or fragment them, which distorts every downstream estimate of prophage content, cargo gene load, and induction potential; anchoring boundary evaluation to viral-like-particle sequencing from induced cultures gave the field a ground truth it had largely lacked. The model is a single fixed checkpoint applied to new genomes without retraining, distributed on Zenodo under CC BY 4.0 with inference code on GitHub, though that repository carries no license file. A second Zenodo deposit, also CC BY 4.0, holds the complete 527,686-sequence training corpus with its splits and labels alongside the training and testing scripts, so the classifier can be retrained rather than only run. Limits remain: contig calling below 1 kb, a range the comparison method VirRep does not cover at all, recovers under half of phage sequences, and GPU inference is unsupported on Apple silicon, leaving CPU as the fallback there.

At a glance

Parameters
650 Million
Released
March 2025
Category
DNA & Gene
Organization
Tsinghua University

Related models

  • PathogenFinder2

    Technical University of Denmark / Technical University of Munich / Bristol Myers Squibb

  • DeepVirus

    Beijing Zhongguancun Academy / Zhongguancun Institute of Artificial Intelligence / Tsinghua University / Peking University / Georgia Institute of Technology

  • BGC-Finder

    Huazhong University of Science and Technology

  • Protein Set Transformer

    University of Wisconsin-Madison / Morgridge Institute for Research / Indian Institute of Technology Madras

  • PLM-ICE

    Michigan Technological University / The Ohio State University / Rochester Institute of Technology

Links

GitHub RepositoryResearch PaperbioRxiv PreprintDatasetDataset

Tags

metagenomicstransfer_learningtransformer

Something wrong?

Much of this page is generated or calculated automatically. Flag anything that looks off and we will re-run it.