sanger-tol · ksenia-krasheninnikova · Oct 31, 2023 · Jun 15, 2023 · Jun 20, 2023 · Jun 22, 2023
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -35,9 +35,10 @@ jobs:
         with:
           version: "${{ matrix.NXF_VER }}"
 
+      - name: Download test data
+        run: |
+          curl https://darwin.cog.sanger.ac.uk/genomeassembly_test_data.tar.gz | tar xzf -
+
       - name: Run pipeline with test data
-        # TODO nf-core: You can customise CI pipeline run tests as required
-        # For example: adding multiple test runs with different parameters
-        # Remember that you can parallelise this by using strategy.matrix
         run: |
-          nextflow run ${GITHUB_WORKSPACE} -profile test,docker --outdir ./results
+          nextflow run ${GITHUB_WORKSPACE} -profile test_github,docker -c conf/hifiasm.config --outdir ./results
diff --git a/assets/test.yaml b/assets/test.yaml
@@ -11,3 +11,7 @@ dataset:
 hic_motif: GATC,GANTC,CTNAG,TTAA
 busco:
   lineage: bacteria_odb10
+mito:
+  species: Caradrina clavipalpis
+  min_length: 15000
+  code: 5
diff --git a/assets/test_github.yaml b/assets/test_github.yaml
@@ -0,0 +1,17 @@
+dataset:
+  id: baUndUnlc1
+  illumina_10X:
+    reads: /home/runner/work/genomeassembly/genomeassembly/Undibacterium_unclassified/genomic_data/baUndUnlc1/10x/
+  pacbio:
+    reads:
+      - reads: /home/runner/work/genomeassembly/genomeassembly/Undibacterium_unclassified/genomic_data/baUndUnlc1/pacbio/fasta/HiFi.reads.fasta
+  HiC:
+    reads:
+      - reads: /home/runner/work/genomeassembly/genomeassembly/Undibacterium_unclassified/genomic_data/baUndUnlc1/hic-arima2/41741_2#7.sub.cram
+hic_motif: GATC,GANTC,CTNAG,TTAA
+busco:
+  lineage: bacteria_odb10
+mito:
+  species: Caradrina clavipalpis
+  min_length: 15000
+  code: 5
diff --git a/assets/test_gsMetZobe1.yaml b/assets/test_gsMetZobe1.yaml
@@ -1,14 +1,18 @@
-samples:
-  - id: gsMetZobe1
-    illumina_10X:
-      reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/gsMetZobe1/10x/
-    pacbio:
-      reads:
-        - reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/gsMetZobe1/pacbio/m64125_200823_145825.ccs.bc1019_BAK8B_OA--bc1019_BAK8B_OA.filtered.fasta.gz
-    HiC:
-      reads:
-        - reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/gsMetZobe1/hic-arima/35528_4#7.cram
-arima_motif: GATC,GANTC,CTNAG,TTAA
+dataset:
+  id: gsMetZobe1
+  illumina_10X:
+    reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/gsMetZobe1/10x/
+  pacbio:
+    reads:
+      - reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/gsMetZobe1/pacbio/m64125_200823_145825.ccs.bc1019_BAK8B_OA--bc1019_BAK8B_OA.filtered.fasta.gz
+  HiC:
+    reads:
+      - reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/gsMetZobe1/hic-arima/35528_4#7.cram
+hic_motif: GATC,GANTC,CTNAG,TTAA
 busco:
   lineages_path: /lustre/scratch123/tol/resources/busco/v5/
   lineage: fungi_odb10
+mito:
+  species: Metschnikowia zobellii
+  min_length: 15000
+  code: 4
diff --git a/assets/test_iyVesGerm1.yaml b/assets/test_iyVesGerm1.yaml
@@ -1,17 +1,18 @@
-samples:
-  - id: iyVesGerm1
-    assembly:
-      primary: /lustre/scratch124/tol/projects/darwin/data/insects/Vespula_germanica/working/iyVesGerm1.hifiasm.20201015/iyVesGerm1.p_ctg.fa.gz
-      haplotigs: /lustre/scratch124/tol/projects/darwin/data/insects/Vespula_germanica/working/iyVesGerm1.hifiasm.20201015/iyVesGerm1.a_ctg.fa.gz
-    illumina_10X:
-      reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/sanger-tol-genomeassembly/test/10X_iyVesGerm1
-    pacbio:
-      reads:
-        - reads: /lustre/scratch124/tol/projects/darwin/data/insects/Vespula_germanica/genomic_data/iyVesGerm1/pacbio/fasta/m64094_200217_145414.ccs.bc1011_BAK8A_OA--bc1011_BAK8A_OA.filtered.fasta.gz
-    HiC:
-      reads:
-        - reads: /lustre/scratch124/tol/projects/darwin/data/insects/Vespula_germanica/genomic_data/iyVesGerm1/hic-arima2/34957_3#2.cram
-      arima_motif: GATC,GANTC,CTNAG,TTAA
-    busco:
-      lineages_path: /lustre/scratch123/tol/resources/busco/v5/
-      lineage: insecta_odb10
+dataset:
+  id: iyVesGerm1
+  illumina_10X:
+    reads: /lustre/scratch124/tol/projects/darwin/users/kk16/development/nextflow/sanger-tol-genomeassembly/test/10X_iyVesGerm1
+  pacbio:
+    reads:
+      - reads: /lustre/scratch124/tol/projects/darwin/data/insects/Vespula_germanica/genomic_data/iyVesGerm1/pacbio/fasta/m64094_200217_145414.ccs.bc1011_BAK8A_OA--bc1011_BAK8A_OA.filtered.fasta.gz
+  HiC:
+    reads:
+      - reads: /lustre/scratch124/tol/projects/darwin/data/insects/Vespula_germanica/genomic_data/iyVesGerm1/hic-arima2/34957_3#2.cram
+hic_motif: GATC,GANTC,CTNAG,TTAA
+busco:
+  lineages_path: /lustre/scratch123/tol/resources/busco/v5/
+  lineage: insecta_odb10
+mito:
+  species: Vespula germanica
+  min_length: 15000
+  code: 5
diff --git a/conf/hifiasm.config b/conf/hifiasm.config
@@ -0,0 +1,18 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Nextflow config file for running minimal tests
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Defines input files and everything required to run a fast and simple pipeline test.
+
+    Use as follows:
+        nextflow run sanger-tol/genomeassembly -profile test,<docker/singularity> --outdir <OUTDIR>
+
+----------------------------------------------------------------------------------------
+*/
+
+process {
+    withName: '.*HIFIASM.*' {
+        // Skip bloom filter
+        ext.args = '--primary -f0'
+    }
+}
diff --git a/conf/modules.config b/conf/modules.config
@@ -172,6 +172,27 @@ process {
     }
     // End of Set up of the raw assembly pipeline
 
+    // Set up of organelles pipeline
+
+    withName: '.*ORGANELLES_READS:MITOHIFI_MITOHIFI' {
+        ext.args2 = '-r'
+        publishDir = [
+            path: { "${params.outdir}/${meta.id}.${params.hifiasm}/mito.reads" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+    }
+
+    withName: '.*ORGANELLES_CONTIGS:MITOHIFI_MITOHIFI' {
+        ext.args2 = '-c'
+        publishDir = [
+            path: { "${params.outdir}/${meta.id}.${params.hifiasm}/mito" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+    }
+    // End of set up of organelles pipeline
+
     // Set up of the purging pipeline
 
     withName: '.*PURGE_DUPS_PRI:MINIMAP2_ALIGN_READS' {

diff --git a/conf/test.config b/conf/test.config
@@ -26,8 +26,10 @@ params {
     hifiasm_hic_on = true
 }
 
-// Set up of the scffolding eipeline
-withName: 'YAHS' {
-    // Skip the initial assembly error correction step
-    ext.args = '-r 1000,2000,5000'
+process {
+    // Set up of the scffolding eipeline
+    withName: 'YAHS' {
+        // Skip the initial assembly error correction step
+        ext.args = '-r 1000,2000,5000'
+    }
 }
diff --git a/conf/test_github.config b/conf/test_github.config
@@ -0,0 +1,36 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Nextflow config file for running minimal tests
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Defines input files and everything required to run a fast and simple pipeline test.
+
+    Use as follows:
+        nextflow run sanger-tol/genomeassembly -profile test,<docker/singularity> --outdir <OUTDIR>
+
+----------------------------------------------------------------------------------------
+*/
+
+params {
+    config_profile_name        = 'Github test profile'
+    config_profile_description = 'Minimal test dataset to check pipeline function'
+
+    max_cpus   = 2
+    max_memory = '6.GB'
+    max_time   = '6.h'
+    // Limit resources so that this can run on GitHub Actions
+
+    // Input data
+    input = "${projectDir}/assets/test_github.yaml"
+    bed_chunks_polishing = 2
+    polishing_on = false
+    hifiasm_hic_on = true
+    organelles_on = false
+}
+
+process {
+    // Set up of the scffolding eipeline
+    withName: 'YAHS' {
+        // Skip the initial assembly error correction step
+        ext.args = '-r 1000,2000,5000'
+    }
+}
diff --git a/modules.json b/modules.json
@@ -130,6 +130,22 @@
                         "installed_by": ["modules"],
                         "patch": "modules/nf-core/minimap2/align/minimap2-align.diff"
                     },
+                    "mitohifi/findmitoreference": {
+                        "branch": "master",
+                        "git_sha": "f52220e84bfc16a8616a5bb3d6f5bc67d601bdce",
+                        "installed_by": ["modules"]
+                    },
+                    "mitohifi/mitohifi": {
+                        "branch": "master",
+                        "git_sha": "02c3c39e3d57bf77183e563598bae2ec1ec11cb1",
+                        "installed_by": ["modules"],
+                        "patch": "modules/nf-core/mitohifi/mitohifi/mitohifi-mitohifi.diff"
+                    },
+                    "multiqc": {
+                        "branch": "master",
+                        "git_sha": "ee80d14721e76e2e079103b8dcd5d57129e584ba",
+                        "installed_by": ["modules"]
+                    },
                     "pretextmap": {
                         "branch": "master",
                         "git_sha": "0f8a77ff00e65eaeebc509b8156eaa983192474b",

diff --git a/modules/local/get_calcuts_params.nf b/modules/local/get_calcuts_params.nf
@@ -4,7 +4,7 @@ process GET_CALCUTS_PARAMS {
     conda "conda-forge::python"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
         'https://depot.galaxyproject.org/singularity/python:3' :
-        'python:3' }"
+        'biocontainers/python:3.9' }"
 
     input:
     tuple val(meta), path(model_fk)

diff --git a/modules/local/keep_seqnames.nf b/modules/local/keep_seqnames.nf
@@ -4,7 +4,7 @@ process KEEP_SEQNAMES {
     conda "conda-forge::coreutils=9.1"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
         'https://depot.galaxyproject.org/singularity/ubuntu:20.04' :
-        'ubuntu:20.04' }"
+        'docker.io/ubuntu:20.04' }"
 
     input:
     tuple val(meta), path(fasta)

diff --git a/modules/nf-core/mitohifi/findmitoreference/main.nf b/modules/nf-core/mitohifi/findmitoreference/main.nf
diff --git a/modules/nf-core/mitohifi/findmitoreference/meta.yml b/modules/nf-core/mitohifi/findmitoreference/meta.yml