Add found_in tag to CNV-calling

genomic-medicine-sweden · Sep 30, 2024 · 1961f37 · 1961f37
1 parent e9f1aea
commit 1961f37
Show file tree

Hide file tree

Showing 5 changed files with 33 additions and 20 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -19,6 +19,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - [#388](https://github.com/genomic-medicine-sweden/nallo/pull/388) - Added longphase as the default phaser
 - [#388](https://github.com/genomic-medicine-sweden/nallo/pull/388) - Added single-sample tbi output to the short variant calling subworkflow
 - [#393](https://github.com/genomic-medicine-sweden/nallo/pull/393) - Added a new `--minimap2_read_mapping_preset` parameter
+- [#403](https://github.com/genomic-medicine-sweden/nallo/pull/403) - Added `FOUND_IN=hificnv` tags to CNV output
 
 ### `Changed`
 
@@ -38,6 +39,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - [#393](https://github.com/genomic-medicine-sweden/nallo/pull/393) - Changed the default minimap2 preset for PacBio data from `map-hifi` to `lr:hqae`
 - [#397](https://github.com/genomic-medicine-sweden/nallo/pull/397) - Changed `pipelines_testdata_base_path` to pin a specific commit
 - [#402](https://github.com/genomic-medicine-sweden/nallo/pull/402) - Updated broken test profile link added in [#397](https://github.com/genomic-medicine-sweden/nallo/pull/397)
+- [#403](https://github.com/genomic-medicine-sweden/nallo/pull/403) - Changed `ADD_FOUND_IN_TAG` process to allow input files to be named the same as output
 
 ### `Removed`
 

diff --git a/conf/modules/cnv.config b/conf/modules/cnv.config
@@ -22,11 +22,15 @@ process {
         publishDir = [
             path: { "${params.outdir}/cnv_calling/hificnv/${meta.id}" },
             mode: params.publish_dir_mode,
-            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename.endsWith('.vcf.gz') ? null : filename }
         ]
     }
 
-    withName: '.*:CNV:TABIX_HIFICNV' {
+    withName: '.*:CNV:ADD_FOUND_IN_TAG' {
+        ext.args = [
+            '--output-type z',
+            '--write-index=tbi'
+        ].join(' ')
         publishDir = [
             path: { "${params.outdir}/cnv_calling/hificnv/${meta.id}" },
             mode: params.publish_dir_mode,

diff --git a/modules/local/add_found_in_tag/main.nf b/modules/local/add_found_in_tag/main.nf
@@ -8,7 +8,7 @@ process ADD_FOUND_IN_TAG {
         'biocontainers/bcftools:1.20--h8b25389_0' }"
 
     input:
-    tuple val(meta), path(vcf), path(index)
+    tuple val(meta), path(vcf, stageAs: "?/*"), path(index, stageAs: "?/*")
     val(variant_caller)
 
     output:
@@ -28,6 +28,7 @@ process ADD_FOUND_IN_TAG {
                     args.contains("--output-type z") || args.contains("-Oz") ? "vcf.gz" :
                     args.contains("--output-type v") || args.contains("-Ov") ? "vcf" :
                     "vcf"
+
     """
     bcftools view \\
         --threads $task.cpus \\

diff --git a/subworkflows/local/cnv.nf b/subworkflows/local/cnv.nf
@@ -1,5 +1,5 @@
+include { ADD_FOUND_IN_TAG             } from '../../modules/local/add_found_in_tag'
 include { HIFICNV                      } from '../../modules/local/pacbio/hificnv'
-include { TABIX_TABIX as TABIX_HIFICNV } from '../../modules/nf-core/tabix/tabix/'
 
 workflow CNV {
 
@@ -42,10 +42,16 @@ workflow CNV {
     HIFICNV(ch_hificnv_in, ch_fasta, ch_exclude_bed)
     ch_versions = ch_versions.mix(HIFICNV.out.versions)
 
-    TABIX_HIFICNV(HIFICNV.out.vcf)
-    ch_versions = ch_versions.mix(TABIX_HIFICNV.out.versions)
+    // Add FOUND_IN=hificnv to VCF
+    ADD_FOUND_IN_TAG (
+        HIFICNV.out.vcf.map { meta, vcf -> [ meta, vcf, [] ] },
+        "hificnv"
+    )
+    ch_versions = ch_versions.mix(ADD_FOUND_IN_TAG.out.versions)
 
     emit:
-    versions = ch_versions                  // channel: [ versions.yml ]
+    vcf = ADD_FOUND_IN_TAG.out.vcf // channel: [ val(meta), path(vcf) ]
+    tbi = ADD_FOUND_IN_TAG.out.tbi // channel: [ val(meta), path(tbi) ]
+    versions = ch_versions         // channel: [ versions.yml ]
 }
 
diff --git a/tests/main.nf.test.snap b/tests/main.nf.test.snap
@@ -30,8 +30,8 @@
             "HG002_Revio.hap2.paf.gz:md5,a99e451e4af96a7ab120cedd49eea618",
             "HG002_Revio.hap2.var.gz:md5,53dc16fc893a529a95067da161b4dad6",
             "HG002_Revio.pair.vcf.gz:md5,a390558ab50ce0dc9efec070b1c70c22",
-            "HG002_Revio.vcf.gz:md5,771f7d9ecbd161dedfbdd053ceff0d52",
-            "HG002_Revio.vcf.gz.tbi:md5,65d6930f79e626a5542091b7a419dc2b",
+            "HG002_Revio.vcf.gz:md5,e790fd716d526b4fb2683a5064590e30",
+            "HG002_Revio.vcf.gz.tbi:md5,930be7eb79286a615f965962f44a3e18",
             "HG002_Revio_modkit_pileup.bed.gz:md5,55496cdcd7e4ccad86adee3cb819da56",
             "HG002_Revio_modkit_pileup.bed.gz.tbi:md5,b217c5e971ecb0b8c874b0038f0d2ae4",
             "fcc3f220bb09d3264f97e7f620779be0",
@@ -52,7 +52,7 @@
             "nf-test": "0.9.0",
             "nextflow": "24.04.4"
         },
-        "timestamp": "2024-09-30T16:14:48.238330482"
+        "timestamp": "2024-09-30T16:41:22.321880678"
     },
     "test profile - multisample": {
         "content": [
@@ -81,8 +81,8 @@
             "HG002_Revio_A.hap2.paf.gz:md5,a99e451e4af96a7ab120cedd49eea618",
             "HG002_Revio_A.hap2.var.gz:md5,53dc16fc893a529a95067da161b4dad6",
             "HG002_Revio_A.pair.vcf.gz:md5,233375a4067362bb6799d5ade46db537",
-            "HG002_Revio_A.vcf.gz:md5,e575fbadcd826243561192310f234c82",
-            "HG002_Revio_A.vcf.gz.tbi:md5,e27fe63e51a2cfaec020ad0a1aec4a7f",
+            "HG002_Revio_A.vcf.gz:md5,901c55b4db621e15fd752e029da2a30b",
+            "HG002_Revio_A.vcf.gz.tbi:md5,eab1874c171f8fa91dfcc948bc1fb872",
             "HG002_Revio_A_modkit_pileup.bed.gz:md5,55496cdcd7e4ccad86adee3cb819da56",
             "HG002_Revio_A_modkit_pileup.bed.gz.tbi:md5,b217c5e971ecb0b8c874b0038f0d2ae4",
             "fcc3f220bb09d3264f97e7f620779be0",
@@ -120,8 +120,8 @@
             "HG002_Revio_B.hap2.paf.gz:md5,14e230db02f0fefcacdd45949743d80c",
             "HG002_Revio_B.hap2.var.gz:md5,5ed54d6881aac6c54245bb8c5ed77d44",
             "HG002_Revio_B.pair.vcf.gz:md5,62c75ce4450a5fde9fdc6e552f5c2cd9",
-            "HG002_Revio_B.vcf.gz:md5,76bf24c26a6310f885a03a64a511098e",
-            "HG002_Revio_B.vcf.gz.tbi:md5,ef6bc66b37b179727ac0c8635d39a8dd",
+            "HG002_Revio_B.vcf.gz:md5,a49a55ab1f5bdd2c7d36bd23da56db0c",
+            "HG002_Revio_B.vcf.gz.tbi:md5,6667ae233f4bdd2df9406eea8b7d2249",
             "HG002_Revio_B_modkit_pileup.bed.gz:md5,c217ccd3e8514a6a5eca0e163dee9f20",
             "HG002_Revio_B_modkit_pileup.bed.gz.tbi:md5,383b647625175e59420c5d8044ba711f",
             "fcc3f220bb09d3264f97e7f620779be0",
@@ -142,7 +142,7 @@
             "nf-test": "0.9.0",
             "nextflow": "24.04.4"
         },
-        "timestamp": "2024-09-30T16:16:20.523890327"
+        "timestamp": "2024-09-30T16:42:53.500697246"
     },
     "test profile - multisample - ont - parallel_alignments 2 - parallel_snv 1": {
         "content": [
@@ -156,8 +156,8 @@
             "HG002_ONT_A.mosdepth.global.dist.txt:md5,5ae0972357f99aa481a0bf12fb9e0b0b",
             "HG002_ONT_A.mosdepth.region.dist.txt:md5,023b1c6aeaf8fa5ededd6b711a5cd012",
             "HG002_ONT_A.mosdepth.summary.txt:md5,c3b664b0983213f73edf3c0d5a0b04a2",
-            "HG002_ONT_A.vcf.gz:md5,4814c13b0127f88bfa7857bf19e17875",
-            "HG002_ONT_A.vcf.gz.tbi:md5,0a5ace7b099fbe88ba3806e7ce83d653",
+            "HG002_ONT_A.vcf.gz:md5,58061e208274d75853c159f4626190d5",
+            "HG002_ONT_A.vcf.gz.tbi:md5,4f03375bf2b6cff5f746936a71676add",
             "HG002_ONT_A_modkit_pileup.bed.gz:md5,e04d51deb9019e67e7ddd7f1bd82a580",
             "HG002_ONT_A_modkit_pileup.bed.gz.tbi:md5,1e198df01c63afea13ef9d04e2d03684",
             "7b8dc04d36dc2290fd031da73e8c8553",
@@ -177,8 +177,8 @@
             "HG002_ONT_B.mosdepth.global.dist.txt:md5,e551c7a9d1a26cecc99d15ac93a4a924",
             "HG002_ONT_B.mosdepth.region.dist.txt:md5,11154f617f447f4c59dc192d9af4249f",
             "HG002_ONT_B.mosdepth.summary.txt:md5,bcb06b29c22cb6ad200f283e72c29653",
-            "HG002_ONT_B.vcf.gz:md5,c4b0cf6c2efebbf746c6b4d8b6c94c57",
-            "HG002_ONT_B.vcf.gz.tbi:md5,de293b6c556d537ccfddf659b138610b",
+            "HG002_ONT_B.vcf.gz:md5,d7a9976c874dae7b62ec3498c0986747",
+            "HG002_ONT_B.vcf.gz.tbi:md5,9dc5e9553e5937ea7b34b973a64076ff",
             "HG002_ONT_B_modkit_pileup.bed.gz:md5,95a96ceddc795c9c5c4a8d9e6f9d3b5b",
             "HG002_ONT_B_modkit_pileup.bed.gz.tbi:md5,8276c6feeac0be16274198831c75023d",
             "7b8dc04d36dc2290fd031da73e8c8553",
@@ -196,6 +196,6 @@
             "nf-test": "0.9.0",
             "nextflow": "24.04.4"
         },
-        "timestamp": "2024-09-30T16:17:44.422130232"
+        "timestamp": "2024-09-30T16:44:18.189450874"
     }
 }