Efficient detection and typing of phage-plasmids

Link to bioRxiv:

https://doi.org/10.1101/2025.08.29.673033

Link to GitHub: 

https://github.com/EpfeiferNutri/Phage-plasmids

Summary:

Phage-plasmids are both temperate phages and plasmids. They facilitate genetic exchanges and recombination between the two and this dual nature complicates their accurate classification.

To address this challenge, we developed tyPPing, a user-friendly computational pipeline designed to detect and type the most common P-P types. tyPPing uses curated protein profiles to identify new P-P candidates with high accuracy. It also assigns confidence scores to its predictions to facilitate the result interpretation.

In our study, we applied tyPPing to a collection of complete and draft bacterial genomes and identified numerous high-confidence P-Ps. Experiments confirmed that selected candidates were active phage-plasmids. Compared to existing methods (such as MM-GRC, vConTACT v2, geNomad) tyPPing demonstrated superior sensitivity and scalability. We also examine the current limitations in detecting rare or poorly characterized P-P types and suggest new strategies for improving future identification efforts.

-  tyPPing_signature_profiles.hmm – P-P HMMs, specific to P-P types and used for protein-to-profile searches by tyPPing.

-  phage.hmm – phage-specific HMM profiles used by MM-GRC.

-  models/ – random forest models trained to detect P-Ps in plasmid datasets (used by MM-GRC).

-  draft_genomes_analysis/ – incomplete genomes from carbapenem-resistant Enterobacteriales (CRE) species. It includes short-read assemblies, long-read assemblies, complete P-P genomes (organized in corresponding subfolders), and prediction tables generated by tyPPing_for_draft_genomes.R.

-  g2g_plot_tables/ – input tables used to generate genome-to-genome comparative plots.

-  tyPPing_criteria_tables/ – Tables how MinProteins and Composition thresholds (required for tyPPing) were chosen.

- PPs_complete_assemblies - Complete sequences of P-Ps detected in draft genomes of CRE species (P-Ps found in PMID: 36154183 and this study)

- Reads_PPs_MMC_induction - Reads (short) from the MMC induction experiments of strain 170D8, 174J8 and 204G7. Also available at the European nucleotide archive under PRJEB92425.

- Draft_genomes_long_reads - Long reads (PacBio) generated by whole genome sequencing of strain 170D8, 174J8 and 204G7. Also available at the European nucleotide archive under PRJEB92425.