Added an option to skip filtering hits from the same species

sanger-tol · Aug 22, 2024 · cd471de · cd471de
1 parent 70f961c
commit cd471de
Show file tree

Hide file tree

Showing 4 changed files with 15 additions and 5 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -15,13 +15,16 @@ The pipeline has now been validated for draft (unpublished) assemblies.
   automatically selecting based on the taxonomy.
 - All parameters are now passed the regular Nextflow way. There is no support
   for the original Yaml configuration files of the Snakemake version.
+- New option `--skip_taxon_filtering` to skip the taxon filtering in blast searches.
+  Mostly relevant for draft assemblies.
 
 ### Parameters
 
-| Old parameter | New parameter    |
-| ------------- | ---------------- |
-| --yaml        |                  |
-|               | --busco_lineages |
+| Old parameter | New parameter          |
+| ------------- | ---------------------- |
+| --yaml        |                        |
+|               | --busco_lineages       |
+|               | --skip_taxon_filtering |
 
 > **NB:** Parameter has been **updated** if both old and new parameter information is present. </br> **NB:** Parameter has been **added** if just the new parameter information is present. </br> **NB:** Parameter has been **removed** if new parameter information isn't present.
 

diff --git a/nextflow.config b/nextflow.config
@@ -33,6 +33,7 @@ params {
     blastp                     = null
     blastx                     = null
     blastn                     = null
+    skip_taxon_filtering       = false
 
     // MultiQC options
     multiqc_config             = null

diff --git a/nextflow_schema.json b/nextflow_schema.json
@@ -32,6 +32,11 @@
                     "description": "Turn on optional genome masking if needed.",
                     "fa_icon": "fas fa-toggle-off"
                 },
+                "skip_taxon_filtering": {
+                    "type": "boolean",
+                    "description": "Skip filtering out hits from the same species in blast* searches.",
+                    "fa_icon": "fas fa-toggle-off"
+                },
                 "fetchngs_samplesheet": {
                     "type": "boolean",
                     "description": "Turn on the conversion from a nf-core/fetchngs samplesheet.",

diff --git a/subworkflows/local/input_check.nf b/subworkflows/local/input_check.nf
@@ -87,9 +87,10 @@ workflow INPUT_CHECK {
 
 
     //
-    // Get the taxon ID
+    // Get the taxon ID if we do taxon filtering in blast* searches
     //
     ch_parsed_csv.taxon_id
+    | map { params.skip_taxon_filtering ? '' : it }
     | first
     | set { ch_taxon_id }