From 557d31dfd2fecbac2a415ddab780e6a3abc87168 Mon Sep 17 00:00:00 2001
From: ljmesi <37740329+ljmesi@users.noreply.github.com>
Date: Thu, 5 May 2022 13:19:10 +0200
Subject: [PATCH] Add parameter for turning on longread host removal

---
 conf/test.config     | 2 +-
 docs/usage.md        | 2 +-
 nextflow.config      | 2 +-
 nextflow_schema.json | 4 +++-
 4 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/conf/test.config b/conf/test.config
index a2464b2..6d04f60 100644
--- a/conf/test.config
+++ b/conf/test.config
@@ -28,7 +28,7 @@ params {
     perform_longread_clip                 = false
     perform_shortread_complexityfilter    = true
     perform_shortread_hostremoval         = true
-    shortread_hostremoval_reference       = 'https://raw.githubusercontent.com/nf-core/test-datasets/modules/data/genomics/homo_sapiens/genome/genome.fasta'
+    perform_longread_hostremoval          = true
     run_kaiju                             = true
     run_kraken2                           = true
     run_malt                              = true
diff --git a/docs/usage.md b/docs/usage.md
index cee2bb6..537b94a 100644
--- a/docs/usage.md
+++ b/docs/usage.md
@@ -191,7 +191,7 @@ You can optionally save the FASTQ output of the run merging with the `--save_com
 
 #### Host Removal
 
-Removal of possible-host reads from FASTQ files prior profiling can be activated with `--perform_shortread_hostremoval`
+Removal of possible-host reads from FASTQ files prior profiling can be activated with `--perform_shortread_hostremoval` or `--perform_longread_hostremoval`.
 
 Similarly to complexity filtering, host-removal can be useful for runtime optimisation and reduction in misclassified reads. It is not always necessary to report classification of reads from a host when you already know the host of the sample, therefore you can gain a run-time and computational advantage by removing these prior typically resource-heavy profiling with more efficient methods. Furthermore, particularly with human samples, you can reduce the number of false positives during profiling that occur due to host-sequence contamination in reference genomes on public databases.
 
diff --git a/nextflow.config b/nextflow.config
index 8c99af2..4ac0c44 100644
--- a/nextflow.config
+++ b/nextflow.config
@@ -82,7 +82,7 @@ params {
 
     // Host Removal
     perform_shortread_hostremoval          = false
-    shortread_hostremoval_reference        = null
+    perform_longread_hostremoval           = false
     shortread_hostremoval_index            = null
     longread_hostremoval_index             = null
     save_hostremoval_index                 = false
diff --git a/nextflow_schema.json b/nextflow_schema.json
index 9e4cc6c..d2eee95 100644
--- a/nextflow_schema.json
+++ b/nextflow_schema.json
@@ -362,7 +362,9 @@
         "perform_shortread_hostremoval": {
             "type": "boolean"
         },
-        "shortread_hostremoval_reference": {
+        "perform_longread_hostremoval": {
+            "type": "boolean"
+        },
             "type": "string",
             "default": "None"
         },