nf-core/funcscan

Before running the pipeline, you will need to create a design file with information about the samples to be scanned by nf-core/funcscan, containing sample name and path/to/your/contigs.fasta. Use this parameter to specify its location. It has to be a comma-separated file with 2 columns, and a header row (sample, fasta). See usage docs.

Set this parameter to your e-mail address to get a summary e-mail with details of the run sent to you when the workflow exits. If set in your user config file (~/.nextflow/config) then you don't need to specify this on the command line for every run.

nf-core/funcscan will split unannotated input contigs into two categories based on this parameter: long, and short.

For both ARG and AMPs, both categories of contigs will be screened. For BGCs, only contigs in the 'long' category will be screened.

This is due to an (approximate) 'biological' minimal that nucleotide sequences need to be to synthesise a valid BGC, as well as to speeding up pipeline runs by screening only useful contigs.

File and sample ID names in results tables will be indicated by _long and _short suffixes.

This flag turns on the taxonomic classification of input nucleotide sequences. The taxonomic annotations should be turned on if the input metagenomes' bacterial sources are unknown, which can help identify the source of the AMP, BGC or ARG hit obtained for laboratory experiments. This flag should be turned off (which is by default) if the input nucleotide sequences represent a single known genome or nf-core/mag was run beforehand. Turning on this flag relatively decreases the pipeline speed and requires >8GB RAM. Due to the size of the resulting table, the final complete summary is in a zipped format.

This flag specifies which tool for taxonomic classification should be activated. At the moment only 'MMseqs2' is incorporated in the pipeline.

Specify a path to a database that is prepared in MMseqs2 format as detailed in the documentation.

Specify which MMseqs2-formatted database to use to classify the input contigs. This can be a nucleotide or amino acid database that includes taxonomic classifications. For example, both GTDB (an amico acid database) and SILVA (a nucleotide database) are supported by MMseqs2. More details can be found in the documentation.\n\n> Modifies tool parameter(s):\n> - mmseqs databases <name>".

This flag saves the temporary files from downloading the database and formatting it in the MMseqs2 format into the output folder. More details can be found in the documentation.\n\n> Modifies tool parameter(s):\n> - mmseqs databases: --remove-tmp-files".

This flag saves the temporary files from creating the taxonomy database and the final tsv file into the output folder. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --remove-tmp-files".

Specify the type of alignment to be carried out between the query database and the reference MMseqs2 database. This can be set to '0' for automatic detection, '1' for amino acid alignment, '2' for translating the inputs and running the alignment on the translated sequences, '3' nucleotide based alignment and '4' for the translated nucleotide sequences alignment. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --search-type".

Specify the taxonomic ranks to include in the taxonomic lineage column in the final .tsv file. For example, 'kingdom,phylum,class,order,family,genus,species'. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --lca-ranks".

This flag specifies whether the taxonomic lineage should be included in the output .tsv file. The taxonomic lineage is obtained from the internal module of mmseqs/taxonomy that infers the last common ancestor to classify the taxonomy. A value of '0' writes no taxonomic lineage, a value of '1' adds a column with the full lineage names prefixed with abbreviation of the lineage level, e.g. k_Prokaryotes;p_Bacteroidetes;c_....;o_....;f_....;g_....;s_...., while a value of '2' adds a column with the full NCBI taxids lineage,e.g. 1324;2345;4546;5345. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --tax-lineage".

This flag specifies the speed and sensitivity of the taxonomic search. It stands for how many kmers should be produced during the preliminary seeding stage. A very fast search requires a low value e.g. '1.0' and a a very sensitive search requires e.g. '7.0'. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --s".

This flag specifies the sensitivity used for prefiltering the query ORF. Before the taxonomy-assigning step, MMseqs2 searches the predicted ORFs against the provided database. This value influences the speed with which the search is carried out. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --orf-filter-s".

This flag specifies the strategy used for assigning the last common ancestor (LCA). MMseqs2 assigns taxonomy based on an accelerated approximation of the 2bLCA protocol and uses the value of '3'. In this mode, the taxonomic assignment is based not only on usual alignment parameters but also considers the taxonomic classification of the LCA. When the value '4' is used the LCA is assigned based on all the equal scoring top hits. If the value '1' is used the LCA assignment is disregarded and the taxonomic assignment is based on usual alignment parameters like E-value and coverage. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --lca-mode".

This flag assigns the mode value with which the weights are computed. The value of '0' stands for uniform weights of taxonomy assignments, the value of '1' uses the minus log E-value and '2' the actual score. More details can be found in the documentation. \n\n> Modifies tool parameter(s):\n> - mmseqs taxonomy: --vote-mode".

Specify a path to a database that is prepared in a BAKTA format.

If you want the pipeline to download the Bakta database for you, you can choose between the full (33.1 GB) and light (1.3 GB) version. The full version is generally recommended for best annotation results, because it contains all of these:

• UPS: unique protein sequences identified via length and MD5 hash digests (100% coverage & 100% sequence identity)
• IPS: identical protein sequences comprising seeds of UniProt's UniRef100 protein sequence clusters
• PSC: protein sequences clusters comprising seeds of UniProt's UniRef90 protein sequence clusters
• PSCC: protein sequences clusters of clusters comprising annotations of UniProt's UniRef50 protein sequence clusters

If download bandwidth, storage, memory, or run duration requirements become an issue, go for the light version (which only contains PSCCs) by modifying the annotation_bakta_db_downloadtype flag.
More details can be found in the documentation

Modifies tool parameter(s):

• BAKTA_DBDOWNLOAD: --type

By default, Bakta's --meta mode is used in the pipeline to improve the gene prediction of highly fragmented metagenomes.

By specifying this parameter Bakta will instead use its default mode that is optimised for singular 'complete' genome sequences.

More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --meta

Specify the minimum contig size that would be annotated by BAKTA.
If run with '--annotation_bakta_compliant', the minimum contig length must be set to 200. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --min-contig-length

Specify the genetic code translation table used for translation of nucleotides to amino acids.
All possible genetic codes (1-25) used for gene annotation can be found here. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --translation-table

Specify the type of bacteria expected in the input dataset for correct annotation of the signal peptide predictions. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --gram

This flag expects contigs that make up complete chromosomes and/or plasmids. By calling it, the user ensured that the contigs are complete replicons. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --complete

This flag specifies that the contig headers should be rewritten. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --keep-contig-headers

The resulting annotations are cleaned up to standardise them to Genbank/ENA/DDJB conventions. CDS without any attributed hits and those without gene symbols or product descriptions different from hypothetical will be marked as 'hypothetical'.
When activated the '--min-contig-length' will be set to 200. More info can be found here.

Modifies tool parameter(s):

• BAKTA: --compliant

This flag activates tRNAscan-SE 2.0 that predicts tRNA genes. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --skip-trna

This flag activates Aragorn that predicts tmRNA genes. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --skip-tmrna
  `

This flag activates Infernal vs. Rfam rRNA covariance models that predicts rRNA genes. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --rrna

This flag activates Infernal vs. Rfam ncRNA covariance models that predicts ncRNA genes.
BAKTA distinguishes between ncRNA genes and (cis-regulatory) regions to enable the distinction of feature overlap detection.
This including distinguishing between ncRNA gene types: sRNA, antisense, ribozyme and antitoxin. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --ncrna

This flag activates Infernal vs. Rfam ncRNA covariance models that predicts ncRNA cis-regulatory regions.
BAKTA distinguishes between ncRNA genes and (cis-regulatory) regions to enable the distinction of feature overlap detection.
This including distinguishing between ncRNA (cis-regulatory) region types: riboswitch, thermoregulator, leader and frameshift element. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --skip-ncrna-region

This flag activates PILER-CR that predicts CRISPR arrays. More details can be found in the documentation.

Modifies tool parameter(s):

• BAKTA: --skip-crispr

This flag skips CDS prediction that is done by PYRODIGAL with which the distinct prediction for complete replicons and uncompleted contigs is done.
For more information on how BAKTA predicts CDS please refer to BAKTA documentation.

Modifies tool parameter(s):

• BAKTA: --skip-cds

This flag activates the search for reference Phytochelatin Synthase genes (PCSs) using hypothetical CDS as seed sequences, then aligns the translated PCSs against up-/downstream-elongated CDS regions. For more info refer to BAKTA documentation.

Modifies tool parameter(s):

• BAKTA: --skip-pseudo

Skip the prediction of sORFs from amino acids stretches as less than 30aa. For more info please refer to BAKTA documentation. All sORF without gene symbols or product descriptions different from hypothetical will be discarded, while only those identified hits exhibiting proper gene symbols or product descriptions different from hypothetical will still be included in the final annotation.

Modifies tool parameter(s):

• BAKTA: --skip-sorf

Activates any gene annotation found within contig assembly gaps. For more info. please refer to BAKTA documentation.

Modifies tool parameter(s):

• BAKTA: --skip-gap

Activates the BAKTA search for oriC/oriT genes by comparing results from Blast+ (generated by cov=0.8, id=0.8) and the MOB-suite of oriT & DoriC oriC/oriV sequences. Annotations of ori regions take into account overlapping Blast+ hits and are conducted based on a majority vote heuristic. Region edges may be fuzzy. For more info please refer to the BAKTA documentation.

Modifies tool parameter(s):

• BAKTA: --skip-ori

Activate this flag to generate genome plots (might be memory-intensive).

Modifies tool parameter(s):

• BAKTA: --skip-plot

By default, Prokka's --metagenome mode is used in the pipeline to improve the gene prediction of highly fragmented metagenomes.

By specifying this parameter Prokka will instead use its default mode that is optimised for singular 'complete' genome sequences.

For more information, please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --metagenome

By default, annotation in Prokka is carried out by alignment to other proteins in its database, or the databases the user provides via the tools --proteins flag. The resulting annotations are then cleaned up to standardise them to Genbank/ENA conventions.
'Vague names' are set to 'hypothetical proteins', 'possible/probable/predicted' are set to 'putative' and 'EC/CPG and locus tag ids' are removed.

By supplying this flag you stop such clean up leaving the original annotation names.

For more information please check Prokka documentation.

This flag suppresses this default behavior of Prokka (which is to perform the cleaning).

Modifies tool parameter(s):

• Prokka: --rawproduct

Specifies the kingdom that the input sample is derived from and/or you wish to screen for

⚠️ Prokka cannot annotate Eukaryotes.

For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --kingdom

Specify the translation table used to annotate the sequences. All possible genetic codes (1-25) used for gene annotation can be found here. This flag is required if the flag --kingdom is assigned.

For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --gcode

Specify the minimum contig lengths to carry out annotations on. The Prokka developers recommend that this should be >= 200 bp, if you plan to submit such annotations to NCBI.

For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --mincontiglen

Specifiy the minimum e-value used for filtering the alignment hits.

For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --evalue

Specify the minimum coverage percent of the annotated genome. This must be set between 0-100.

For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --coverage

Allow transfer RNA (trRNA) to overlap coding sequences (CDS). Transfer RNAs are short stretches of nucleotide sequences that link mRNA and the amino acid sequence of proteins. Their presence helps in the annotation of the sequences, because each trRNA can only be attached to one type of amino acid.

For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --cdsrnaolap

Activates RNAmmer instead of the Prokka default Barrnap for rRNA prediction during the annotation process. RNAmmer classifies ribosomal RNA genes in genome sequences by using two levels of Hidden Markov Models. Barrnap uses the nhmmer tool that includes HMMER 3.1 for HMM searching in RNA:DNA style.

For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --rnammer

Force the contig headers to conform to the Genbank/ENA/DDJB contig header standards. This is activated in combination with --centre [X] when contig headers supplied by the user are non-conforming and therefore need to be renamed before Prokka can start annotation. This flag activates --genes --mincontiglen 200. For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --compliant

For every CDS annotated, this flag adds the gene that encodes for that CDS region. For more information please check Prokka documentation.

Modifies tool parameter(s):

• Prokka: --addgenes

This parameter allows prokka to retain the original contig names by activating PROKKA's --force flag. If this parameter is set to false it activates PROKKA's flags --locus-tag PROKKA --centre CENTER so the locus tags (contig names) will be PROKKA_# and the center tag will be CENTER. By default PROKKA changes contig headers to avoid errors that might rise due to long contig headers, so this must be turned on if the user has short contig names that should be retained by PROKKA.

Modifies tool parameter(s):

• Prokka: --locus-tag PROKKA --centre CENTER
• Prokka: --force

By default Prodigal runs in 'single genome' mode that requires sequence lengths to be equal or longer than 20000 characters.

However, more fragmented reads from MAGs often result in contigs shorter than this. Therefore, nf-core/funcscan will run with the meta mode by default, but providing this parameter allows to override this and run in single genome mode again.

For more information check Prodigal documentation.

Modifies tool parameter(s):
-PRODIGAL: -p

Suppresses partial genes from being on contig edge, resulting in closed ends. Should only be activated for genomes where it is sure the first and last bases of the sequence(s) do not fall inside a gene. Run together with -p normal (former -p single) .

For more information check Prodigal documentation.

Modifies tool parameter(s):

• PRODIGAL: -c

Specifies which translation table should be used for seqeunce annotation. All possible genetic code translation tables can be found here. The default is set at 11, which is used for standard Bacteria/Archeae.

For more information check Prodigal documentation.

Modifies tool parameter(s):

• PRODIGAL: -g

Forces PRODIGAL to a full scan for motifs rather than activating the Shine-Dalgarno RBS finder, the default scanner for PRODIGAL to train for motifs.

For more information check Prodigal documentation.

Modifies tool parameter(s):

• PRODIGAL: -n

By default Pyrodigal runs in 'single genome' mode that requires sequence lengths to be equal or longer than 20000 characters.

However, more fragmented reads from MAGs often result in contigs shorter than this. Therefore, nf-core/funcscan will run with the meta mode by default, but providing this parameter allows to override this and run in single genome mode again.

For more information check Pyrodigal documentation.

Modifies tool parameter(s):
-PYRODIGAL: -p

Suppresses partial genes from being on contig edge, resulting in closed ends. Should only be activated for genomes where it is sure the first and last bases of the sequence(s) do not fall inside a gene. Run together with -p single .

For more information check Pyrodigal documentation.

Modifies tool parameter(s):

• PYRODIGAL: -c

Specifies which translation table should be used for seqeunce annotation. All possible genetic code translation tables can be found here. The default is set at 11, which is used for standard Bacteria/Archeae.

For more information check Pyrodigal documentation.

Modifies tool parameter(s):

• PYRODIGAL: -g

Forces Pyrodigal to a full scan for motifs rather than activating the Shine-Dalgarno RBS finder, the default scanner for Pyrodigal to train for motifs.

For more information check Pyrodigal documentation.

Modifies tool parameter(s):

• PYRODIGAL: -n

While nf-core/funcscan can download databases for you, often these are very large and can significantly slow-down pipeline runtime if the databases have to be downloaded every run.

Specifying --save_databases while save the pipeline-downloaded databases in your results directory. This applies to: AMRFinderPlus, antiSMASH, BAKTA, CARD, DeepARG, DeepBGC, and DRAMP.

You can then move the resulting directories/files to a central cache directory of your choice for re-use in the future.

If you do not specify these flags, the database files will remain in your work/ directory and will be deleted if cleanup = true is specified in your config, or if you run nextflow clean.

AMPir uses a supervised statistical machine learning approach to predict AMPs. It incorporates two support vector machine classification models, "precursor" and "mature".

The precursor module is better for predicted proteins from a translated transcriptome or translated gene models. The alternative model (mature) is best suited for AMP sequences after post-translational processing, typically from direct proteomic sequencing.

More information can be found in the AMPir documentation.

Modifies tool parameter(s):

• AMPir: model =

Filters result for minimum protein length.
Note that amino acid sequences that are shorter than 10 amino acids long and/or contain anything other than the standard 20 amino acids are not evaluated and will contain an NA as their prob_AMP value

More information can be found in the AMPir documentation.

Modifies tool parameter(s):

• AMPir parameter: min_length in the calculate_features() function

HMMSearch performs biosequence analysis using profile hidden Markov Models.
The models are specified in.hmm files that are specified with this parameter

e.g.

--amp_hmmsearch_models '/<path>/<to>/<models>/*.hmm'  

You must wrap the path in quotes if you use a wildcard, to ensure Nextflow expansion not bash!

For more information check HMMER documentation.

Save a multiple alignment of all significant hits (those satisfying inclusion thresholds) to a file

For more information check HMMER documentation.

Modifies tool parameter(s):

• HMMsearch: -A

Save a simple tabular (space-delimited) file summarizing the per-target output, with one data line per homologous target sequence found.

For more information check HMMER documentation.

Modifies tool parameter(s)

• HMMsearch: --tblout

Save a simple tabular (space-delimited) file summarizing the per-domain output, with one data line per homologous domain detected in a query sequence for each homologous model.

For more information check HMMER documentation.

Modifies tool parameter(s):

• HMMsearch: --domtblout

AMPcombi uses the 'general AMPs' dataset of the (DRAMP database)[http://dramp.cpu-bioinfor.org/downloads/] for taxonomic classification. If you have a local version of it, you can provide the path to the folder containing the reference database files:

1. a fasta file with a .fasta file extension
2. the corresponding table with with functional and taxonomic classifications in .tsv file extension.

For more information check AMPcombi documentation.

Specify the minimum probability an AMP hit must have to be retained in the final output file. Anything below this threshold will be removed.

For more information check AMPcombi documentation.

Modifies tool parameter(s):

• AMPCOMBI: --cutoff

Specify the path to a local version of the ARMFinderPlus database. If no input is given, the pipeline will download the database for you.

See the nf-core/funcscan usage documentation for more information.

Specify the minimum percentage amino-acid identity to reference protein or nucleotide identity for nucleotide reference must have if a BLAST alignment (based on methods: BLAST or PARTIAL) was detected, otherwise NA.

If you specify -1, this means use a curated threshold if it exists and 0.9 otherwise.

Setting this value to something other than -1 will override any curated similarity cutoffs. For BLAST: alignment is > 90% of length and > 90% identity to a protein in the AMRFinderPlus database. For PARTIAL: alignment is > 50% of length, but < 90% of length and > 90% identity to the reference, and does not end at a contig boundary.

For more information check AMRFinderPlus documentation.

Modifies tool parameter(s):

• AMRFinderPlus: --ident_min

Minimum proportion of reference gene covered for a BLAST-based hit analysis if a BLAST alignment was detected, otherwise NA.

For BLAST-based hit analysis: alignment is > 90% of length and > 90% identity to a protein in the AMRFinderPlus database or for PARTIAL: alignment is > 50% of length, but < 90% of length and > 90% identity to the reference, and does not end at a contig boundary.

For more information check AMRFinderPlus documentation.

Modifies tool parameter(s):

• AMRFinderPlus: --coverage_min

NCBI genetic code for translated BLAST. Number from 1 to 33 to represent the translation table used for BLASTX.

See translation table for more details on which table to use.

For more information check AMRFinderPlus documentation.

Modifies tool parameter(s):

• AMRFinderPlus: --translation_table

Provide results from "Plus" genes in the output files.

Mostly the plus genes are an expanded set of genes that are of interest in pathogens. This set includes stress response (biocide, metal, and heat resistance), virulence factors, some antigens, and porins. These "plus" proteins have primarily been added to the database with curated BLAST cutoffs, and are generally identified by BLAST searches. Some of these may not be acquired genes or mutations, but may be intrinsic in some organisms. See AMRFinderPlus database for more details.

Modifies tool parameter(s):

• AMRFinderPlus: --plus

Prepend a column containing an identifier for this run of AMRFinderPlus. For example this can be used to add a sample name column to the AMRFinderPlus results. If set to true, the --name <identifier> is the sample name.

Modifies tool parameter(s):

• AMRFinderPlus: --name

Specify the path to a local version of the DeepARG database (see the pipelines' usage documentation). If no input is given, the module will download the database for you, however this is not recommended, as the database is large and this will take time.

The DeepARG tool itself does not report explicit the database version it uses. We assume the latest version (as downloaded by the tool's database download module), however if you supply a different database, you must supply the version with this parameter for use with the downstream hAMRonization tool.

The version number must be without any leading v etc.

Specify which model to use: short sequences for reads (SS), or long sequences for genes (LS). In the vast majority of cases we recommend using the LS model when using funcscan

For more information check DeepARG documentation.

Modifies tool parameter(s):

• DeepARG: --model

Sets the minimum probability cutoff below which hits are discarded.

For more information check DeepARG documentation.

Modifies tool parameter(s):

• DeepARG: --min-prob

Sets the cutoff value for Evalue below which hits are discarded

For more information check DeepARG documentation.

Modifies tool parameter(s):

• DeepARG: --arg-alignment-evalue

Sets the value for Identity cutoff for sequence alignment

For more information check DeepARG documentation.

Modifies tool parameter(s):

• DeepARG: --arg-alignment-identity

Sets the value for the allowed alignment read overlap.

For more information check DeepARG documentation.

Modifies tool parameter(s):

• DeepARG: --arg-alignment-overlap

Sets the value of minimum number of alignments per entry for DIAMOND.

For more information check DeepARG documentation.

Modifies tool parameter(s):

• DeepARG: --arg-num-alignments-per-entry

Specify via a comma separated list any of the hmm-models of the pre-defined models:
- Class A beta-lactamases: class_a
- Subclass B1 and B2 beta-lactamases: class_b_1_2
- Subclass B3 beta-lactamases: class_b_3
- Class C beta-lactamases: class_c - Class D beta-lactamases: class_d_1, class_d_2 - qnr:qnr - Tetracycline resistance genestet_efflux, tet_rpg, tet_enzyme`

For more information check fARGene documentation.

For example: --arg_fargenemodel 'class_a,qnr,tet_enzyme'

Modifies tool parameter(s):

• fARGene: --hmm-model

fARGene generates many additional temporary files which in most cases won't be useful and thus by default are not saved to the pipeline's result directory.

By specifying this parameter, the directories tmpdir/, hmmsearchresults/ and spades_assemblies/ will be also saved in the output directory for closer inspection by the user, if necessary.

The threshold score for a sequence to be classified as a (almost) complete gene. If not pre-assigned, it is assigned by the hmm_model used based on the trade-off between sensitivity and specificity.

For more details see code documentation.

Modifies tool parameter(s):

• fARGene: --score

The minimum length of a predicted ORF retrieved from annotating the nucleotide sequences. By default the pipeline assigns this to 90% of the assigned hmm_model sequence length.

For more information check fARGene documentation.

Modifies tool parameter(s):

• fARGene: --min-orf-length

By default, pipeline uses prodigal/prokka for the prediction of ORFs from nucleotide sequences. Another option is the NCBI ORFfinder tool that is built into fARGene, the use of which is activated by this flag.

For more information check fARGene documentation.

Modifies tool parameter(s):

• fARGene: --orf-finder

The translation format that transeq should use for amino acid annotation from the nucleotide sequences. More sequence formats can be found in transeq 'input sequence formats'.

For more information check fARGene documentation.

Modifies tool parameter(s):

• fARGene: --translation-format

You can pre-download the CARD database to your machine and pass the path of it to this parameter.

See the pipeline documentation for details on how to download this.

Modifies tool parameter(s):

• RGI_CARDANNOTATION: --input

When activated, this flag saves the .json file in the RGI output directory. The .json file contains the ARG predictions in a format that can be can be uploaded to the CARD website for visualization. See RGI documentation for more details. By default, the .json file is generated in the working directory but not saved in the results directory to save disk space (.json file is quite large and not required downstream in the pipeline).

RGI generates many additional temporary files which in most cases won't be useful so by default are not saved.

By specifying this parameter, the files including temp in the name will be also saved in the output directory for closer inspection by the user, if necessary.

Specifies the alignment tool to be used. By default RGI runs BLAST and this is also set as default in the nf-core/funcscan pipeline. Using this flag the user can activate the alignment by DIAMOND again.

For more information check RGI documentation.

Modifies tool parameter(s):

• RGI_MAIN: --alignment_tool

When activated it includes 'Loose' hits (a.k.a. Discovery) in addition to strict and perfect hits. All 'Loose' matches of 95% identity or better are automatically listed as 'Strict', regardless of alignment length (RGI v. <6.0.0). This behaviour can be overrun by using the --include_nudge flag. The 'Loose' algorithm works outside of the detection model cut-offs to provide detection of new, emergent threats and more distant homologs of AMR genes, but will also catalog homologous sequences and spurious partial matches that may not have a role in AMR.

For more information check RGI documentation.

Modifies tool parameter(s):

• RGI_MAIN: --include_loose

This flag suppresses the default behaviour of RGI with --include_loose, which lists all 'Loose' matches of >= 95% identity as 'Strict', regardless of alignment length. With this strict and perfect labels are added. This is discontinued in future versions of RGI.

For more information check RGI documentation.

Modifies tool parameter(s):

• RGI_MAIN: --include_nudge

This flag should be used only when the contigs are of poor quality (e.g. short) to predict partial genes.

For more information check RGI documentation.

Modifies tool parameter(s):

• RGI_MAIN: --low_quality

This flag is used to specify the data type used as input to RGI. By default this is set as 'NA', which makes no assumptions on input data.

For more information check RGI documentation.

Modifies tool parameter(s):

• RGI_MAIN: --data

Modifies tool parameter:

• RGI_MAIN: --split_prodigal_jobs

Specifies which database to use from dedicated list of databases available by ABRicate.

For more information check ABRicate documentation.

Modifies tool parameter(s):

• ABRicate: --db

Specifies the minimum percent identity used to classify an ARG hit using BLAST alignment.

For more information check ABRicate documentation.

Modifies tool parameter(s):

• ABRicate: --minid

Specifies the minimum coverage of the nucleotide sequence to be assigned an ARG hit using BLAST alignment. In the ABRicate matrix, an absent gene is assigned (.) and if present, it is assigned the estimated coverage (#).

For more information check ABRicate documentation.

Modifies tool parameter(s):

• ABRicate: --mincov

It is recommend to pre-download the antiSMASH databases to your machine and pass the path of it to this parameter, as this can take a long time to download - particularly when running lots of pipeline runs.

See the pipeline documentation for details on how to download this. If running with docker or singularity, please also check --bgc_antismash_installationdirectory for important information.

This is required when running with docker and singularity (not required for conda), due to attempted 'modifications' of files during database checks in the installation directory, something that cannot be done in immutable docker/singularity containers.

Therefore, a local installation directory needs to be mounted (including all modified files from the downloading step) to the container as a workaround.

This specifies the minimum length that a contig must have for the contig to be screened by antiSMASH.

For more information see the antiSMASH documentation.

This will only apply to samples that are screened with antiSMASH (i.e., those samples that have not been removed by --bgc_antismash_sampleminlength).

You may wish to increase this value compared to that of --bgc_antismash_sampleminlength, in cases where you wish to screen higher-quality (i.e., longer) contigs, or speed up runs by not screening lower quality/less informative contigs.

Modifies tool parameter(s):

• antiSMASH: --minlength

Compare identified clusters against a database of antiSMASH-predicted clusters using the clusterblast algorithm.

For more information see the antiSMASH documentation.

Modifies tool parameter(s):

• antiSMASH: --cb-general

This will turn on comparing identified clusters against known gene clusters from the MIBiG database using the clusterblast algorithm.

MIBiG is a curated datbase of experimentally characterised gene clusters and with rich associated metadata.

For more information see the antiSMASH documentation.

Modifies tool parameter(s):

• antiSMASH: --cb-knownclusters

Turn on additional screening for operons involved in the biosynthesis of early secondary metabolites components using the clusterblast algorithm.

For more information see the antiSMASH documentation.

Modifies tool parameter(s):

• antiSMASH: --cb-subclusters

Turn on comparison of detected genes against the MIBiG database using the ClusterCompare algorithm - an alternative to clusterblast.

Note there will not be a dedicated ClusterCompare output in the antiSMASH results directory, but is present in the HTML.

For more information see the antiSMASH documentation.

Modifies tool parameter(s):

• antiSMASH: --cc-mibig

Turning this on will activate the generation of additional functional and phyogenetic analysis of genes, via comparison against databases of protein orthologs.

For more information see the antiSMASH documentation.

Modifies tool parameter(s):

• antiSMASH: --cb-smcog-trees

Defines which level of strictness to use for HMM-based cluster detection.

These correspond to screening of different groups of 'how well-defined' clusters are. For example, loose will include screening for 'poorly defined' clusters (e.g. saccharides), relaxed for partially present clusters (e.g. certain types of NRPS), whereas strict will screen for well-defined clusters such as Ketosynthases.

You can see the rules for the levels of strictness here.

For more information see the antiSMASH documentation.

Modifies tool parameter(s):

• antiSMASH: --hmmdetection-strictness

This specifies which set of secondary metabolites to screen for, based on the taxon type the secondary metabolites are from.

This will run different pipelines depending on whether the input sequences are from bacteria or fungi.

For more information see the antiSMASH documentation.

Modifies tool parameter(s):

• antiSMASH: --taxon

The DeepBGC score threshold for extracting BGC regions from Pfam sequences based on average protein-wise value. This is a prediction score that the domain is a part of a BGC.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --score

By default DeepBGC's Prodigal runs in 'single genome' mode that requires sequence lengths to be equal or longer than 20000 characters.

However, more fragmented reads from MAGs often result in contigs shorter than this. Therefore, nf-core/funcscan will run with the meta mode by default, but providing this parameter allows to override this and run in single genome mode again.

For more information check Prodigal documentation.

Modifies tool parameter(s)

• DeepBGC: --prodigal-meta-mode

Merge detected BGCs within given number of proteins.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --merge-max-protein-gap

Merge detected BGCs within given number of proteins.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --merge-max-nucl-gap

Minimum length a BGC must have (in bp) to be reported as detected.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --min-nucl

Minimum number of proteins in a BGC must have to be reported as 'detected'.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --min-proteins

Minimum number of domains a BGC must have to be reported as 'detected'.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --min-domains

Minimum number of biosynthetic protein domains a BGC must have to be reported as 'detected'. This is based on antiSMASH definitions.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --min-bio-domains

DeepBGC classification score threshold for assigning classes to BGCs.

For more information see the DeepBGC documentation.

Modifies tool parameter(s)

• DeepBGC: --classifier-score

Enable unknown region masking to prevent genes from stretching across unknown nucleotides during ORF detection based on P(y)rodigal.

For more information see the GECCO documentation.

Modifies tool parameter(s):

• GECCO: --mask

Specify the number of consecutive genes a hit must have to be considered a part of a possible BGC region during BGC extraction.

For more information see the GECCO documentation.

Modifies tool parameter(s):

• GECCO: --cds

The p-value cutoff for protein domains to be included.

For more information see the GECCO documentation.

Modifies tool parameter(s):

• GECCO: --pfilter

Specify the minimum probability a predicted gene must have to be considered a part of a BGC during BGC extraction.

Reducing this value may increase number and length of hits, but will reduce the accuracy of the predictions.

For more information see the GECCO documentation.

Modifies tool parameter(s):

• GECCO: --threshold

The minimum number of annotated genes that must separate a possible BGC cluster from the edge. Edge clusters will still be included if they are longer. A lower number will increase the number of false positives on small contigs. Used during BGC extraction.

For more information see the GECCO documentation.

Modifies tool parameter(s):

• GECCO: --edge-distance

HMMSearch performs biosequence analysis using profile hidden Markov Models.
The models are specified in.hmm files that are specified with this parameter

e.g.

--bgc_hmmsearch_models '/<path>/<to>/<models>/*.hmm'  

You must wrap the path in quotes if you use a wildcard, to ensure Nextflow expansion not bash!

For more information check HMMER documentation.

Save a multiple alignment of all significant hits (those satisfying inclusion thresholds) to a file

For more information check HMMER documentation.

Modifies tool parameter(s):

• HMMsearch: -A

Save a simple tabular (space-delimited) file summarizing the per-target output, with one data line per homologous target sequence found.

For more information check HMMER documentation.

Modifies tool parameter(s)

• HMMsearch: --tblout

Save a simple tabular (space-delimited) file summarizing the per-domain output, with one data line per homologous domain detected in a query sequence for each homologous model.

For more information check HMMER documentation.

Modifies tool parameter(s)

• HMMsearch:--domtblout

Specifies which summary report format to generate with hamronize summarize: tsv, json or interactive (html)

Modifies tool parameter(s)

• HMMsearch: -t, --summary_type