add module parameters

conf/modules.config

@@ -15,15 +15,15 @@ process {
     publishDir = [
         path: { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" },
         mode: params.publish_dir_mode,
-        saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        saveAs: { filename -> filename.equals('versions.yml') ? null : filename },
     ]
 
     withName: MULTIQC {
         ext.args   = { params.multiqc_title ? "--title \"${params.multiqc_title}\"" : '' }
         publishDir = [
             path: { "${params.outdir}/multiqc" },
             mode: params.publish_dir_mode,
-            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename },
         ]
     }
 
@@ -33,7 +33,7 @@ process {
             path: { "${params.outdir}/cat" },
             mode: params.publish_dir_mode,
             saveAs: { filename -> filename.equals('versions.yml') ? null : filename },
-            enabled: params.save_concatenated_fastas
+            enabled: params.save_concatenated_fastas,
         ]
     }
 
@@ -43,15 +43,39 @@ process {
             path: { "${params.outdir}/cat" },
             mode: params.publish_dir_mode,
             saveAs: { filename -> filename.equals('versions.yml') ? null : filename },
-            enabled: params.save_concatenated_fastas
+            enabled: params.save_concatenated_fastas,
         ]
     }
 
+    withName: BRACKEN_BUILD {
+        ext.args = { "${params.bracken_build_params}" }
+    }
+
+    withName: CENTRIFUGE_BUILD {
+        ext.args = { "${params.centrifuge_build_params}" }
+    }
+
+    withName: DIAMOND_MAKEDB {
+        ext.args = { "${params.diamond_build_params}" }
+    }
+
     withName: GANON_BUILD {
-        ext.args = { "--verbose" }
+        ext.args = { "${params.ganon_build_params}" }
+    }
+
+    withName: KAIJU_MKFMI {
+        ext.args = { "${params.kaiju_build_params}" }
+    }
+
+    withName: KRAKEN2_BUILD {
+        ext.args = { "${params.kraken2_build_params}" }
+    }
+
+    withName: KRAKENUNIQ_BUILD {
+        ext.args = { "${params.krakenuniq_build_params} --jellyfish-bin type -P -a jellyfish)" }
     }
 
     withName: MALT_BUILD {
-        ext.args = { "--sequenceType ${params.malt_sequencetype}" }
+        ext.args = { "${params.malt_build_params}" }
     }
 }

nextflow.config

@@ -59,15 +59,22 @@ params {
 
     // tool specific options
     build_bracken                = false
-    build_centrifuge             = false
     build_diamond                = false
     build_ganon                  = false
     build_kaiju                  = false
+    build_malt                   = false
+    build_centrifuge             = false
     build_kraken2                = false
     kraken2_keepintermediate     = false
     build_krakenuniq             = false
-    build_malt                   = false
-    malt_sequencetype            = "DNA"
+    bracken_build_params         = ''
+    centrifuge_build_params      = ''
+    diamond_build_params         = ''
+    ganon_build_params           = ''
+    kaiju_build_params           = ''
+    kraken2_build_params         = ''
+    krakenuniq_build_params      = ''
+    malt_build_params            = "--sequenceType DNA"
 }
 
 // Load base.config by default for all pipelines

nextflow_schema.json

@@ -113,31 +113,67 @@
                     "description": "Turn on extending of Kraken2 database to include Bracken files. Requires nucleotide FASTA File input.",
                     "help_text": "Bracken2 databases are simply just a Kraken2 database with two additional files.\n\nNote however this requires a Kraken2 database _with_ intermediate files still in it, thus can result in large database directories."
                 },
+                "bracken_build_params": {
+                    "type": "string",
+                    "description": "Specify parameters being given to bracken build. Must be wrapped in single and double quotes: --bracken_build_params \"'--your_param'\"",
+                    "help_text": "See [Bracken documentation](https://github.com/jenniferlu717/Bracken?tab=readme-ov-file#step-1-generate-the-bracken-database-file-databasexmerskmer_distrib).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "fa_icon": "fas fa-users-cog"
+                },
                 "build_centrifuge": {
                     "type": "boolean",
                     "description": "Turn on building of Centrifuge database. Requires nucleotide FASTA file input.",
                     "fa_icon": "fas fa-toggle-on"
                 },
+                "centrifuge_build_params": {
+                    "type": "string",
+                    "description": "Specify parameters being given to centrifuge-build. Must be wrapped in single and double quotes: --centrifuge_build_params \"'--your_param'\"",
+                    "help_text": "See [Centrifuge documentation](https://github.com/DaehwanKimLab/centrifuge/blob/master/MANUAL.markdown#database-download-and-index-building).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "fa_icon": "fas fa-users-cog"
+                },
                 "build_diamond": {
                     "type": "boolean",
                     "fa_icon": "fas fa-toggle-on",
                     "description": "Turn on building of DIAMOND database. Requires amino-acid FASTA file input."
                 },
+                "diamond_build_params": {
+                    "type": "string",
+                    "description": "Specify parameters being given to diamond makedb. Must be wrapped in single and double quotes: --diamond_build_params \"'--your_param'\"",
+                    "help_text": "See [diamond documentation](https://github.com/bbuchfink/diamond/wiki/3.-Command-line-options#makedb-options).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "fa_icon": "fas fa-users-cog"
+                },
                 "build_ganon": {
                     "type": "boolean",
                     "description": "Turn on building of ganon database. Requires nucleotide FASTA file input.",
                     "fa_icon": "fas fa-toggle-on"
                 },
+                "ganon_build_params": {
+                    "type": "string",
+                    "description": "Specify parameters being given to ganon buildcustom. Must be wrapped in single and double quotes: --ganon_build_params \"'--your_param'\"",
+                    "help_text": "See [ganon documentation](https://pirovc.github.io/ganon/custom_databases/).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "fa_icon": "fas fa-users-cog"
+                },
                 "build_kaiju": {
                     "type": "boolean",
                     "description": "Turn on building of Kaiju database. Requires amino-acid FASTA file input.",
                     "fa_icon": "fas fa-toggle-on"
                 },
+                "kaiju_build_params": {
+                    "type": "string",
+                    "description": "Specify parameters being given to kaiju-mkbwt. Must be wrapped in single and double quotes: --kaiju_build_params \"'--your_param'\"",
+                    "help_text": "See [Kaiju documentation](https://github.com/bioinformatics-centre/kaiju/tree/master#custom-database).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "fa_icon": "fas fa-users-cog"
+                },
                 "build_kraken2": {
                     "type": "boolean",
                     "description": "Turn on building of Kraken2 database. Requires nucleotide FASTA file input.",
                     "fa_icon": "fas fa-toggle-on"
                 },
+                "kraken2_build_params": {
+                    "type": "string",
+                    "description": "Specify parameters being given to kraken2 build. Must be wrapped in single and double quotes: --kraken2_build_params \"'--your_param'\"",
+                    "help_text": "See [Kraken2 documentation](https://github.com/DerrickWood/kraken2/blob/master/docs/MANUAL.markdown#custom-databases).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "fa_icon": "fas fa-users-cog"
+                },
                 "kraken2_keepintermediate": {
                     "type": "boolean",
                     "fa_icon": "fas fa-save",
@@ -148,18 +184,23 @@
                     "fa_icon": "fas fa-toggle-on",
                     "description": "Turn on building of KrakenUniq database. Requires nucleotide FASTA file input."
                 },
+                "krakenuniq_build_params": {
+                    "type": "string",
+                    "description": "Specify parameters being given to krakenuniq build. Must be wrapped in single and double quotes: --krakenuniq_build_params \"'--your_param'\"",
+                    "help_text": "See [KrakenUniq documentation](https://github.com/fbreitwieser/krakenuniq?tab=readme-ov-file#database-building).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "fa_icon": "fas fa-users-cog"
+                },
                 "build_malt": {
                     "type": "boolean",
                     "fa_icon": "fas fa-toggle-on",
                     "description": "Turn on building of MALT database. Requires nucleotide FASTA file input."
                 },
-                "malt_sequencetype": {
+                "malt_build_params": {
                     "type": "string",
-                    "default": "DNA",
-                    "description": "Specify type of input sequence being given to MALT",
-                    "enum": ["DNA", "Protein"],
-                    "help_text": "Use to specify whether the reference sequences are DNA or Protein sequences. (For RNA sequences, use the DNA setting) - from [MALT manual](https://software-ab.cs.uni-tuebingen.de/download/malt/).\n\n> Modifies tool(s) parameter(s)\n> - malt-build: `--sequenceType` ",
-                    "fa_icon": "fas fa-dna"
+                    "description": "Specify parameters given to malt-build. Must include --sequenceType DNA or Protein and be wrapped in double and single quotes: --malt_build_params \"'--sequenceType DNA --your_param'\"",
+                    "help_text": "At least the type of reference sequences should be specified. (For RNA sequences, use the DNA setting) - from [MALT manual](https://software-ab.cs.uni-tuebingen.de/download/malt/).\n\nParameters must be wrapped in _both_ single and then double quotes outside these, to ensure the parameters are interpreted as a string for internal use rather than interpreted by the terminal as a pipeline-level parameter.",
+                    "default": "--sequenceType DNA",
+                    "fa_icon": "fas fa-users-cog"
                 }
             },
             "fa_icon": "fas fa-database"

subworkflows/local/utils_nfcore_createtaxdb_pipeline/main.nf

@@ -43,7 +43,7 @@ workflow PIPELINE_INITIALISATION {
         version,
         true,
         outdir,
-        workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1
+        workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1,
     )
 
     //
@@ -52,7 +52,7 @@ workflow PIPELINE_INITIALISATION {
     UTILS_NFSCHEMA_PLUGIN(
         workflow,
         validate_params,
-        null
+        null,
     )
 
     //
@@ -62,6 +62,11 @@ workflow PIPELINE_INITIALISATION {
         nextflow_cli_args
     )
 
+    //
+    // Custom validation for pipeline parameters
+    //
+    validateInputParameters()
+
     //
     // Create channel from input file provided through params.input
     //
@@ -104,7 +109,7 @@ workflow PIPELINE_COMPLETION {
                 plaintext_email,
                 outdir,
                 monochrome_logs,
-                multiqc_report.toList()
+                multiqc_report.toList(),
             )
         }
 
@@ -159,6 +164,10 @@ def validateInputParameters() {
     if (params.build_krakenuniq && [!params.malt_mapdb].any()) {
         error('[nf-core/createtaxdb] Supplied --build_malt, but missing: --malt_mapdb (all are mandatory for MALT)')
     }
+
+    if (params.build_malt && !(params.malt_build_params.contains('--sequenceType DNA') || params.malt_build_params.contains('--sequenceType Protein'))) {
+        error('[nf-core/createtaxdb] Supplied --build_malt, but --malt_build_params must contain at a minimum malt-build parameters --sequenceType DNA or --sequenceType Protein')
+    }
 }
 
 //
@@ -193,7 +202,7 @@ def toolCitationText() {
         params.build_krakenuniq ? "KrakenUniq (Breitwieser et al. 2018)," : "",
         params.build_malt ? "MALT (Vågene et al. 2018)," : "",
         "and MultiQC (Ewels et al. 2016)",
-        "."
+        ".",
     ].join(' ').trim()
 
     return citation_text
@@ -212,7 +221,7 @@ def toolBibliographyText() {
         params.build_kraken2 ? '<li>Wood, D. E., Lu, J., & Langmead, B. (2019). Improved metagenomic analysis with Kraken 2. Genome Biology, 20(1), 257.  <a href="https://doi.org/10.1186/s13059-019-1891-0">10.1186/s13059-019-1891-0</a></li>' : "",
         params.build_krakenuniq ? '<li>Breitwieser, F. P., Baker, D. N., & Salzberg, S. L. (2018). KrakenUniq: confident and fast metagenomics classification using unique k-mer counts. Genome Biology, 19(1), 198.  <a href="https://doi.org/10.1186/s13059-018-1568-0">10.1186/s13059-018-1568-0</a></li>' : "",
         params.build_malt ? '<li>Vågene, Å. J., Herbig, A., Campana, M. G., Robles García, N. M., Warinner, C., Sabin, S., Spyrou, M. A., Andrades Valtueña, A., Huson, D., Tuross, N., Bos, K. I., & Krause, J. (2018). Salmonella enterica genomes from victims of a major sixteenth-century epidemic in Mexico. Nature Ecology & Evolution, 2(3), 520–528.  <a href="https://doi.org/10.1038/s41559-017-0446-6">10.1038/s41559-017-0446-6</a></li>' : "",
-        '<li>Ewels, P., Magnusson, M., Lundin, S., & Käller, M. (2016). MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics , 32(19), 3047–3048. doi: /10.1093/bioinformatics/btw354</li>'
+        '<li>Ewels, P., Magnusson, M., Lundin, S., & Käller, M. (2016). MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics , 32(19), 3047–3048. doi: /10.1093/bioinformatics/btw354</li>',
     ].join(' ').trim()
 
     return reference_text