RdRpCATCH

RNA-dependent RNA polymerase Collaborative Analysis Tool with Collections of pHMMs

RdRpCATCH is collaborative effort to combine various publicly available RNA virus RNA-dependent RNA polymerase pHMM databases in one tool to facilitate their detection in (meta-)transcriptomics data.

RdRpCATCH is written in Python and uses the pyHMMER3 library to perform pHMM searches. In addition, the tool scans each sequence (aa or nt) in the input file with the selected databases and provides the best hit (hit with the highest bitscore across all databases) as output. In addition, RdRpCATCH provides information about the number of profiles that were positive for each sequence across all pHMM databases, and taxonomic information based on the MMseqs2 easy-taxonomy and search modules against a custom RefSeq Riboviria database.

Our web server is live! https://rdrpcatch.bioinformatics.nl

Version 0.0.8 -> 1.0.0 Changelog

Added --default-hmmsearch-params to use HMMER default hmmsearch thresholds (E=10.0, domE=10.0, incE=0.01, incdomE=0.01, automatic Z) overriding any user-specified thresholds and ignoring --zvalue.
Added --alt-mmseqs-tax-db to optionally use an alternative MMseqs2 taxonomy database, either by name (under mmseqs_dbs) or by full path to a pressed MMseqs2 database.
Improved CLI validation for numeric parameters and genetic code values, with clearer error messages for invalid combinations (e.g. --db-options none without --custom-dbs).
Added shorthand aliases Olendraite_fam and Olendraite_gen in --db-options as equivalents to TSA_Olendraite_fam and TSA_Olendraite_gen.
Added –extended-output flag to CLI for optionally keeping additional HMM score columns in the output.

Version 0.0.7 -> 0.0.8 Changelog

Added support for custom pHMM databases. See the Setting up custom pHMM databases section for more information.
All specified flags use ‘-‘ instead of ‘_’ (e.g. --db-dir instead of --db_dir).
Fixed issue with specifying the Lucaprot_HMM and Zayed_HMM databases in the --db-options argument.
Command rdrpcatch download renamed as rdrpcatch databases for clarity, as it now supports adding custom pHMM databases to the RdRpCATCH databases. This is facilitated by the --add-custom-db argument.
Added none option to the --db-options argument to search only against custom databases.

** The tool has been modified to use rolypoly code/approaches **

$images\rdrpcatch_illustration.png$

Supported databases

NeoRdRp ¹ : 1182 pHMMs
NeoRdRp2 ²: 19394 pHMMs
RVMT ³: 710 pHMMs
RdRp-Scan ⁴ : 68 pHMMs
TSA_Olendraite_fam ⁵: 77 pHMMs
TSA_Olendraite_gen ⁶ : 341 pHMMs
LucaProt_HMM⁷ : 754 pHMMs
Zayed_HMM⁸ : 2489 pHMMs

Sakaguchi, S. et al. (2022) ‘NeoRdRp: A comprehensive dataset for identifying RNA-dependent RNA polymerases of various RNA viruses from metatranscriptomic data’, Microbes and Environments, 37(3). doi:10.1264/jsme2.me22001
Sakaguchi, S., Nakano, T. and Nakagawa, S. (2024) ‘Neordrp2 with improved seed data, annotations, and scoring’, Frontiers in Virology, 4. doi:10.3389/fviro.2024.1378695
Neri, U. et al. (2022) ‘Expansion of the global RNA virome reveals diverse clades of bacteriophages’, Cell, 185(21). doi:10.1016/j.cell.2022.08.023
Charon, J. et al. (2022) ‘RDRP-Scan: A bioinformatic resource to identify and annotate divergent RNA viruses in metagenomic sequence data’, Virus Evolution, 8(2). doi:10.1093/ve/veac082
Olendraite, I., Brown, K. and Firth, A.E. (2023) ‘Identification of RNA virus–derived rdrp sequences in publicly available transcriptomic data sets’, Molecular Biology and Evolution, 40(4). doi:10.1093/molbev/msad060
Olendraite, I. (2021) ‘Mining diverse and novel RNA viruses in transcriptomic datasets’, Apollo. Available at: https://www.repository.cam.ac.uk/items/1fabebd2-429b-45c9-b6eb-41d27d0a90c2
Hou, X. and He, Y. et al. (2024) ‘Using artificial intelligence to document the hidden RNA virosphere’, Cell, 187(24). doi:10.1016/j.cell.2024.09.027
Zayed, A. A., et al. (2022) ‘Cryptic and abundant marine viruses at the evolutionary origins of Earth’s RNA virome.’ Science, 376(6589), 156–162. doi:10.1126/science.abm5847

Installation

Prerequisites

For the installation process, conda is required. If you don’t have conda installed, you can find instructions on how to https://docs.conda.io/projects/conda/en/latest/user-guide/install/index.html
Mamba is a faster alternative to conda. If you have it installed, you can use it instead of conda.

Installation steps

Update: Version 1.0.1 is available via Bioconda! Please contact us if you have any issues with installation.

The package is available as a bioconda package. You can install it using the following command:

conda create rdrpcatch -c bioconda rdrpcatch

Alternatively, you can install RdRpCATCH from python package index (PyPI) using pip. This requires the installation of the dependencies manually. The dependencies are:

mmseqs2
seqkit

The dependencies can be installed using conda or mamba. Follow these steps:

Create a new conda environment and install the dependencies:

conda create -n rdrpcatch python=3.12
conda activate rdrpcatch
pip install rdrpcatch
conda install -c bioconda mmseqs2==17.b804f seqkit==2.10.0

Activate the environment and download the RdRpCATCH databases:

conda activate rdrpcatch
rdrpcatch databases --destination-dir path/to/store/databases

Note 1: The databases are large files and may take some time to download (~ 3 GB).
Note 2: The databases are stored in the specified directory, and the path is required to run RdRpCATCH.
Note 3: If you encounter an SSL error while downloading, please try again. The error seems to appear sporadically during testing, and a simple re-initiation of the downloading process seems to fix it.
Note 4: If the SSL error persists, or there is any other error related to downloading the databases, please let us know by raising an issue. In this case, you can download the pre-compiled databases manually via the zenodo repository
10.5281/zenodo.15463729 Save the repository to a local directory of your preferance, and then use the following command:
```
tar -xvf rdrpcatch_dbs.tar
```

Usage

RdRpCATCH can be used as a CLI tool as follows:

# make sure the conda environment is activated
# conda activate rdrpcatch

# scan the input fasta file with the selected databases
rdrpcatch scan -i path/to/input.fasta -o path/to/output_dir -db-dir path/to/database

Input description

The input file can be one or more nucleotide or protein sequences in multi-fasta format. The output directory is where the results will be stored. We recommend specifying the type of the sequence in the command line, An optional argument --seq-type (nuc or prot) can be used to specify if the input fasta file sequences are nucleotide or amino acid.

Setting up custom pHMM databases

It is possible to use custom pHMM databases with RdRpCATCH. As a prerequisite, you need to install the RdRpCATCH databases using the rdrpcatch databases command as described above, to a directory of your choice.

The custom databases should be formatted as follows:

First create a directory and give it a descriptive name, e.g. my_custom_rdrp_database. Important: The name should not contain comma , characters.
Inside the directory put your custom pHMM HMMER pressed database. You can use the hmmpress command of HMMER to create the pressed database from your custom HMM file. This creates a set of files with the same name as the original HMM file, but with different extensions (e.g. .h3f, .h3i, .h3m, .h3p). The directory should contain all these files. Please refer to the HMMER manual for more information on how to create a pressed database from an HMM file. (http://eddylab.org/software/hmmer/Userguide.pdf)
Next you can add the directory to the custom databases that are readable by RdRpCATCH. This can be done by using the rdrpcatch databases command as follows:

rdrpcatch databases --add-custom-db path/to/my_custom_rdrp_database --destination-dir path/that/contains/rdrpcatch/databases

This will add the custom database to the list of databases that can be used with RdRpCATCH.
The custom database can then be used with the rdrpcatch scan command by specifying the --custom-dbs argument as follows:

rdrpcatch scan -i path/to/input.fasta -o path/to/output_dir -db-dir path/to/database --custom-dbs custom_database_name

The custom_database_name should be the name of the directory that contains the custom pHMM files, without the path.
For example, if the custom database is stored in path/to/my_custom_rdrp_database, you would use --custom-dbs my_custom_rdrp_database in the command line.
You can add multiple custom databases by installing them in the same way and specifying them by separating them with commas, e.g. --custom-dbs my_custom_rdrp_database,another_custom_database.
The custom databases can be used in combination with the pre-compiled databases provided by RdRpCATCH. To do this, you can specify the --db_options argument with the names of the pre-compiled databases you want to use, and specify the custom databases with the --custom-dbs argument.
For example, if you want to use the NeoRdRp and RVMT databases along with your custom database, you would use the following command:

rdrpcatch scan -i path/to/input.fasta -o path/to/output_dir -db-dir path/to/database --db-options NeoRdRp,RVMT --custom-dbs my_custom_rdrp_database

Note: By default, RdRpCATCH will search against all pre-compiled databases if no --db_options argument is specified. If you want to use only the custom databases, you can specify --db_options none to avoid searching against the pre-compiled databases.

Commands

The following two commands are available in RdRpCATCH:

rdrpcatch scan
rdrpcatch databases

rdrpcatch databases:

Command to download pre-compiled databases from Zenodo and to set up custom databases. If the databases are already downloaded in the specified directory , the command will check for updates and download the latest version if available.

Argument	Short Flag	Type	Description
`--destination_dir`	`-dest`	PATH	Path to the directory to download HMM databases. [required]
`--concept-doi`	``	TEXT	Zenodo Concept DOI for database repository
`--help`	``		Show help message and exit
`--add-custom-db`	``	PATH	Path to the directory containing custom pHMM files to add to the RdRpCATCH databases.

rdrpcatch scan:

Search a given input using selected RdRp databases.

Argument	Short Flag	Type	Description
`--input`	`-i`	FILE	Path to the input FASTA file. [required]
`--output`	`-o`	DIRECTORY	Path to the output directory. [required]
`--db-dir`	`-db-dir`	PATH	Path to the directory containing RdRpCATCH databases. [required]
`--seq-type`	`-seq-type`	TEXT	Type of sequence to search against: (prot,nuc). If omitted, the type will be auto-detected.
`--gen-code`	`-gen-code`	INTEGER	Genetic code to use for translation. (default: 1). Only a subset of NCBI genetic codes supported by `seqkit translate` is accepted (see CLI help for the list).
`--length-thr`	`-length-thr`	INTEGER	Minimum length threshold for seqkit seq. (default: 400)
`--db-options`	`-dbs`	TEXT	Comma-separated list of pre-installed databases to search against. Valid options: RVMT, NeoRdRp, NeoRdRp.2.1, TSA_Olendraite_fam (or Olendraite_fam), TSA_Olendraite_gen (or Olendraite_gen), RDRP-scan, Lucaprot_HMM, Zayed_HMM, all, none.
`--custom-dbs`		PATH	Comma-separated list of custom databases to search against. Valid options: names of the directories that the custom databases are stored in.
`--alt-mmseqs-tax-db`	`-altmmdb`	TEXT	Optional alternative MMseqs2 taxonomy database. Can be a database name under `mmseqs_dbs` (resolved via `fetch_mmseqs_db_path`) or a path to a pressed MMseqs2 DB (directory or base filename). Defaults to `mmseqs_refseq_riboviria_20250211`.
`--evalue`	`-e`	FLOAT	E-value threshold for HMMsearch. (default: 1e-5). Overridden when `--default-hmmsearch-params` is used.
`--incevalue`	`-incE`	FLOAT	Inclusion E-value threshold for HMMsearch. (default: 1e-5). Overridden when `--default-hmmsearch-params` is used.
`--domevalue`	`-domE`	FLOAT	Domain E-value threshold for HMMsearch. (default: 1e-5). Overridden when `--default-hmmsearch-params` is used.
`--incdomevalue`	`-incdomE`	FLOAT	Inclusion domain E-value threshold for HMMsearch. (default: 1e-5). Overridden when `--default-hmmsearch-params` is used.
`--zvalue`	`-z`	INTEGER	Number of sequences to search against. (default: 1000000). Ignored when `--default-hmmsearch-params` is used (automatic Z).
`--default-hmmsearch-params`		FLAG	Use HMMER default `hmmsearch` thresholds (E=10.0, domE=10.0, incE=0.01, incdomE=0.01, automatic Z), overriding `--evalue/--incevalue/--domevalue/--incdomevalue` and ignoring `--zvalue`.
`--cpus`	`-cpus`	INTEGER	Number of CPUs to use for HMMsearch. (default: 1)
`--bundle`	`-bundle`		Bundle the output files into a single archive. (default: False)
`--keep-tmp`	`-keep-tmp`		Keep the temporary files generated during the analysis. (default: False)
`--overwrite`	`-overwrite`	FLAG	Force overwrite of existing output directory. (default: False)
`--verbose`	`-v`	FLAG	Print verbose output.
`--extended-output`		FLAG	Keep additional HMM score columns (`Best_hit_norm_bitscore_profile`, `Best_hit_norm_bitscore_sequence`, `Best_hit_ID_score`) in the output. (default: False)

Output files

rdrpcatch scan will create a folder with the following structure:

Output	Description
`{prefix}_rdrpcatch_output_annotated.tsv`	A tab-separated file containing the results of the RdRpCATCH analysis.
`{prefix}_rdrpcatch_fasta`	A directory containing the sequences that were identified as RdRp sequences.
`{prefix}_rdrpcatch_plots`	A directory containing the plots generated during the analysis.
`{prefix}_gff_files`	A directory containing the GFF files generated during the analysis. (For now only based on protein sequences)
`tmp`	A directory containing temporary files generated during the analysis. (Only available if the -keep_tmp flag is used )

Output table fields

A summary of the results is stored in the {prefix}_rdrpcatch_output_annotated.tsv file, which contains the following fields: | Field | Description | |——-|———————————————————————————————————————| | Sequence_name | The name of the sequence. | | Translated_sequence_name (frame) | The name of the translated sequence and the frame of the RdRp sequence. | | Sequence_length(AA) | The length of the RdRp sequence in amino acids. | | Total_databases_that_the_sequence_was_detected(No_of_Profiles) | The name of databases and the number of profiles that the RdRp sequence was detected by. | | Best_hit_Database | The database with the best hit. | | Best_hit_profile_name | The name of the profile with the best hit. | | Best_hit_profile_length | The length of the profile with the best hit. | | Best_hit_e-value | The e-value of the best hit. | | Best_hit_bitscore | The bitscore of the best hit. | | RdRp_from(AA) | The start position of the RdRp sequence, in relation to the amino acid sequence. | | RdRp_to(AA) | The end position of the RdRp sequence, in relation to the amino acid sequence. | | Best_hit_profile_coverage | The fraction of the profile that was covered by the RdRp sequence. | | Best_hit_sequence_coverage | The fraction of the sequence that was covered by the RdRp sequence. (Based on aminoacid sequence) | | MMseqs_Taxonomy_2bLCA | The taxonomy of the RdRp sequence based on MMseqs2 easy-taxonomy module against a custom RefSeq Riboviria database. | | MMseqs_TopHit_accession | The accession of the top hit in the RefSeq Riboviria database. | | MMseqs_TopHit_fident | The fraction of identical matches of the top hit in the RefSeq Riboviria database. | | MMseqs_TopHit_alnlen | The alignment length of the top hit in the RefSeq Riboviria database. | | MMseqs_TopHit_eval | The e-value of the top hit in the RefSeq Riboviria database. | | MMseqs_TopHit_bitscore | The bitscore of the top hit in the RefSeq Riboviria database. | | MMseqs_TopHit_qcov | The query coverage of the top hit in the RefSeq Riboviria database. | | MMseqs_TopHit_lineage | The lineage of the top hit in the RefSeq Riboviria database. |

Citations

Our preprint is out! RdRpCATCH: A unified resource for RNA virus discovery using viral RNA-dependent RNA polymerase profile Hidden Markov models https://doi.org/10.64898/2026.02.05.703936

A precompiled version of the used databases is available at Zenodo DOI: 10.5281/zenodo.14358348.
If you use RdRpCATCH, please cite the underlying third party databases :

Acknowledgements

RdRpCATCH is a collaborative effort and we would like to thank all the authors and developers of the underlying databases.

Contact

Dimitris Karapliafis (dimitris.karapliafis@wur.nl), potentially via slack/teams or an issue in the main repo.

Contributing

We welcome pull requests and suggestions via raising issues, emails or Slack

Licence

MIT