Phage and virus biology · Software
RdRpCATCH
Identifies RNA virus RNA-dependent RNA polymerase sequences using aggregated profile hidden Markov models.
What RdRpCATCH is
RdRpCATCH (RNA-dependent RNA polymerase Collaborative Analysis Tool with Collections of pHMMs) is a command-line and web-based bioinformatics tool designed to detect RNA viruses across transcriptomic and metatranscriptomic datasets. It was created by Dimitris Karapliafis and an international collaborative group of researchers.
Detection of novel or divergent RNA viruses often depends on profiling RNA-dependent RNA polymerase (RdRp), the sole universally conserved gene among RNA viruses. RdRpCATCH addresses profile fragmentation by consolidating eight publicly available, independently developed profile hidden Markov model (pHMM) collections into a single workflow.
The software analyses nucleotide or amino acid sequences against selected or custom profile databases, returns the best-scoring hit per query sequence, reports cross-database profile detections, and infers taxonomic lineage using an MMseqs2 search against a custom RefSeq Riboviria reference.
How it works
RdRpCATCH is implemented in Python and incorporates approaches from the rolypoly framework. It uses seqkit for sequence manipulation, such as sequence translation and length filtering (default minimum length of 400), and pyHMMER3 to perform sequence scans across HMMER-pressed profile collections. Queries can be provided as amino acid or nucleotide multi-FASTA files, with automatic sequence-type detection and user-configurable genetic translation codes.
For each input sequence, profile searches evaluate domain and inclusion E-values. Across all chosen databases, RdRpCATCH selects the top hit based on the maximum bitscore and tracks the total number of profile hits per database. Sequences are then assigned taxonomy using the MMseqs2 easy-taxonomy and search modules queried against a default custom RefSeq Riboviria database (or an alternative specified database). The pipeline outputs an annotated TSV summary, matching FASTA files, visual plots, and GFF annotation files for protein inputs.
Key facts
- Core language
- Python (version 3.12 compatible)
- Dependencies
- pyHMMER3, MMseqs2, seqkit
- Input formats
- Multi-FASTA (nucleotide or amino acid sequences)
- Output formats
- TSV, FASTA, GFF, plots
- Bundled collections
- NeoRdRp (1,182 pHMMs), NeoRdRp2 (19,394 pHMMs), RVMT (710 pHMMs), RdRp-Scan (68 pHMMs), TSA_Olendraite_fam (77 pHMMs), TSA_Olendraite_gen (341 pHMMs), LucaProt_HMM (754 pHMMs), Zayed_HMM (2,489 pHMMs)
- Default taxonomy database
- mmseqs_refseq_riboviria_20250211
What you can use it for
- Metatranscriptomic RNA virus discovery
- Scanning assembled metatranscriptome contigs in multi-FASTA format to discover divergent and novel RNA viruses across environmental or clinical samples.
- RdRp domain annotation
- Mapping coordinates, amino acid lengths, coverage, and bitscores for conserved polymerase domains in newly sequenced or translated proteins.
- Benchmarking profile collections
- Evaluating and comparing detection rates across individual RdRp profile sets by selecting specific databases or measuring the number of positive profiles per sequence.
- Riboviria taxonomic classification
- Assigning taxonomic lineages to candidate RdRp sequences using MMseqs2 2bLCA searches against a RefSeq Riboviria database.
Run it here
Ask the research agent and it runs RdRpCATCH as a job on a dedicated machine, then reads the results back to you. Nothing to install, and it keeps running if you close the page.
Try asking
“Search these contigs for RNA virus polymerases with RdRpCATCH.”
Strengths and limits
Good at
- Accommodates both raw nucleotide and translated amino acid FASTA inputs with customisable translation tables.
- Permits seamless inclusion of user-created, HMMER-pressed custom pHMM databases alongside or instead of bundled collections.
- Offers fast pHMM search execution via the pyHMMER3 library without requiring external HMMER binary calls.
Watch out for
- Requires substantial disk space and bandwidth to retrieve the pre-compiled database bundle, which is approximately 3 GB compressed.
- GFF annotation file generation is presently supported only when input sequences are provided as protein sequences.
- Requires external installation of the binary tools MMseqs2 and seqkit when installed via pip rather than Bioconda.
Questions
Can I run RdRpCATCH online without installing it?
Ask the research agent and it runs RdRpCATCH as a job on a dedicated machine, then reads the results back to you. Nothing to install, and it keeps running if you close the page.
What licence is RdRpCATCH under?
MIT (code and the Zenodo database bundle, 10.5281/zenodo.14358348). The terms are published at https://github.com/dimitris-karapliafis/RdRpCATCH.
What input file formats does RdRpCATCH accept?
It accepts multi-FASTA files containing either nucleotide or amino acid sequences. Sequence type can be set explicitly using `--seq-type` or left to the tool to auto-detect.
Can I use custom pHMM databases with RdRpCATCH?
Yes. Custom databases can be prepared using HMMER's `hmmpress`, registered with the `rdrpcatch databases --add-custom-db` command, and specified during scanning via `--custom-dbs`.
Which RdRp pHMM collections are bundled with the tool?
It provides access to eight collections: NeoRdRp, NeoRdRp2, RVMT, RdRp-Scan, TSA_Olendraite_fam, TSA_Olendraite_gen, LucaProt_HMM, and Zayed_HMM.
How does the tool handle taxonomy assignment?
It uses MMseqs2 easy-taxonomy and search modules to assign 2bLCA taxonomy and report top-hit alignments against a custom RefSeq Riboviria database, or an alternative user-specified MMseqs2 database.
How are overlapping profile matches summarised in the results?
The primary TSV output reports the best hit based on the highest bitscore across all searched databases, alongside details on how many total profiles and databases detected the sequence.
How to cite it
Dimitris Karapliafis, Uri Neri, Ingrida Olendraite, Justine Charon, Shoichi Sakaguchi, Xin Hou, Dick de Ridder, Mark P Zwart, Anne Kupczok, RdRpCATCH: a unified resource for RNA virus discovery using viral RNA-dependent RNA polymerase profile Hidden Markov models, NAR Genomics and Bioinformatics, Volume 8, Issue 3, September 2026, lqag076, https://doi.org/10.1093/nargab/lqag076
Written from RdRpCATCH's own README and project page. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.