1 implement alignment with spaceranger

assets/samplesheet_test.csv

@@ -1,3 +1,3 @@
-sample,fastq_1,fastq_2
-Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L001_R1_001.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L001_R2_001.fastq.gz
-Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L002_R1_001.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L002_R2_001.fastq.gz
+sample,fastq_1,fastq_2,image,cytaimage,slide,area,darkimage,colorizedimage,manual_alignment,slidefile
+Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L001_R1_001.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L001_R2_001.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_tissue_image.btf,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_image.tif,H1-84QJZFR,D1,,,,
+Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L002_R1_001.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_S1_L002_R2_001.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_tissue_image.btf,https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_image.tif,H1-84QJZFR,D1,,,,

assets/samplesheet_test_long.csv

@@ -0,0 +1,2 @@
+sample,fastq_1,fastq_2,image,cytaimage,slide,area,darkimage,colorizedimage,manual_alignment,slidefile
+Visium_HD_Human_Colon_Cancer,https://raw.githubusercontent.com/nfdata-omics/test-datasets/spatialomics/Visium_HD_Human_Colon_Cancer/Visium_HD_Human_Colon_Cancer_chr21_sub40000_P2_S2_L001_R1_001.fastq.gz,https://raw.githubusercontent.com/nfdata-omics/test-datasets/spatialomics/Visium_HD_Human_Colon_Cancer/Visium_HD_Human_Colon_Cancer_chr21_sub40000_P2_S2_L001_R2_001.fastq.gz,,https://raw.githubusercontent.com/nfdata-omics/test-datasets/spatialomics/Visium_HD_Human_Colon_Cancer/Visium_HD_Human_Colon_Cancer_P2_image.tif,H1-VM2JXXK,A1,,,,

assets/schema_input.json

@@ -26,6 +26,48 @@
                 "exists": true,
                 "pattern": "^([\\S\\s]*\\/)?[^\\s\\/]+\\.f(ast)?q\\.gz$",
                 "errorMessage": "FastQ file for reads 2 cannot contain spaces and must have extension '.fq.gz' or '.fastq.gz'"
+            },
+            "image": {
+                "type": "string",
+                "format": "file-path",
+                "exists": true,
+                "meta": ["image"]
+            },
+            "cytaimage": {
+                "type": "string",
+                "format": "file-path",
+                "exists": true,
+                "meta": ["cytaimage"]
+            },
+            "slide": {
+                "type": "string",
+                "meta": ["slide"]
+            },
+            "area": {
+                "type": "string",
+                "meta": ["area"]
+            },
+            "darkimage": {
+                "type": "string",
+                "format": "file-path",
+                "exists": true,
+                "meta": ["darkimage"]
+            },
+            "colorizedimage": {
+                "type": "string",
+                "format": "file-path",
+                "exists": true,
+                "meta": ["colorizedimage"]
+            },
+            "manual_alignment": {
+                "type": "string",
+                "meta": ["alignment"]
+            },
+            "slidefile": {
+                "type": "string",
+                "format": "file-path",
+                "exists": true,
+                "meta": ["slidefile"]
             }
         },
         "required": ["sample", "fastq_1"]

conf/modules.config

@@ -18,10 +18,104 @@ process {
         saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
     ]
 
+    withName: 'GUNZIP_FASTA' {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
+    withName: 'GUNZIP_GTF' {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
+    withName: 'GUNZIP_GFF' {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
+    withName: 'UNTAR_SPACERANGER_REF' {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
+    withName: 'GFFREAD' {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
+    withName: SPACERANGER_MKGTF {
+        publishDir = [
+            path: "${params.outdir}/reference/mkgtf",
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+        ext.args = "--attribute=gene_biotype:protein_coding \
+                    --attribute=gene_biotype:lincRNA \
+                    --attribute=gene_biotype:antisense \
+                    --attribute=gene_biotype:IG_LV_gene \
+                    --attribute=gene_biotype:IG_V_gene \
+                    --attribute=gene_biotype:IG_V_pseudogene \
+                    --attribute=gene_biotype:IG_D_gene \
+                    --attribute=gene_biotype:IG_J_gene \
+                    --attribute=gene_biotype:IG_J_pseudogene \
+                    --attribute=gene_biotype:IG_C_gene \
+                    --attribute=gene_biotype:IG_C_pseudogene \
+                    --attribute=gene_biotype:TR_V_gene \
+                    --attribute=gene_biotype:TR_V_pseudogene \
+                    --attribute=gene_biotype:TR_D_gene \
+                    --attribute=gene_biotype:TR_J_gene \
+                    --attribute=gene_biotype:TR_J_pseudogene \
+                    --attribute=gene_biotype:TR_C_gene"
+    }
+
+    withName: SPACERANGER_MKREF {
+        publishDir = [
+            path: "${params.outdir}/reference/mkref",
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+    }
+
+    withName: 'CAT_FASTQ' {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
+    withName: 'FQ_LINT' {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
     withName: FASTQC {
         ext.args = '--quiet'
     }
 
+    withName: SPACERANGER_COUNT {
+        publishDir = [
+            [
+                path: { "${params.outdir}/count/" },
+                mode: params.publish_dir_mode,
+                pattern: "outs/web_summary.html",
+                saveAs: { "${meta.id}_web_summary.html" }
+            ],
+            [
+                path: { "${params.outdir}/count/${meta.id}" },
+                mode: params.publish_dir_mode,
+                pattern: "outs/**",
+                saveAs: { filename -> filename.contains('/') ? filename.substring(filename.indexOf('/') + 1) : filename }
+            ]
+        ]
+        ext.args   = "--create-bam true"
+        time   = { 240.h * task.attempt }
+    }
+
     withName: 'MULTIQC' {
         ext.args   = { params.multiqc_title ? "--title \"$params.multiqc_title\"" : '' }
         publishDir = [

conf/test.config

@@ -12,8 +12,8 @@
 
 process {
     resourceLimits = [
-        cpus: 2,
-        memory: '15.GB',
+        cpus: 4,
+        memory: '12.GB',
         time: '1.h'
     ]
 }
@@ -22,8 +22,9 @@ params {
     config_profile_name        = 'Test profile'
     config_profile_description = 'Minimal test dataset to check pipeline function'
 
-    // Input and output
-    input  = "${projectDir}/assets/samplesheet_test.csv"
-    fasta  = "https://raw.githubusercontent.com/nf-core/test-datasets/references/references/GRCh38_chr21/GRCh38_chr21.fa"
-    gtf    = "https://raw.githubusercontent.com/nf-core/test-datasets/references/references/GRCh38_chr21/GRCh38_chr21.gtf"
+    // Input
+    input             = "${projectDir}/assets/samplesheet_test.csv"
+    probeset          = "https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/human-lung-cancer-post-xenium_hd_ffpe/Visium_HD_Human_Lung_Cancer_HD_Only_Experiment2_probe_set.csv"
+    spaceranger_index = "https://raw.githubusercontent.com/nf-core/test-datasets/spatialvi/testdata/GRCh38.tar.gz"
+
 }

conf/test_long.config

@@ -0,0 +1,32 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    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 nfdata-omics/spatialomics -profile test_long,<docker/singularity> --outdir <OUTDIR>
+
+----------------------------------------------------------------------------------------
+*/
+
+process {
+    resourceLimits = [
+        cpus: 2,
+        memory: '15.GB',
+        time: '1.h'
+    ]
+}
+
+params {
+    config_profile_name        = 'Long Test profile'
+    config_profile_description = 'Minimal test dataset to check pipeline function, with full chromosome and ref building'
+
+    // Input
+    input             = "${projectDir}/assets/samplesheet_test_long.csv"
+    probeset          = "https://raw.githubusercontent.com/nfdata-omics/test-datasets/spatialomics/Visium_HD_Human_Colon_Cancer/GSM8594568_P2CRC_probe_set.csv"
+    fasta             = "https://raw.githubusercontent.com/nfdata-omics/test-datasets/refs/chr21-cellranger/chr21.fa.gz"
+    gtf               = "https://raw.githubusercontent.com/nfdata-omics/test-datasets/refs/chr21-cellranger/chr21.gtf.gz"
+    reference_name    = "GRCh38"
+
+}

main.nf

@@ -25,7 +25,7 @@ include { getGenomeAttribute      } from './subworkflows/local/utils_nfcore_spat
 */
 
 params.fasta             = getGenomeAttribute('fasta')
-params.gft               = getGenomeAttribute('gtf')
+params.gtf               = getGenomeAttribute('gtf')
 params.spaceranger_index = getGenomeAttribute('spaceranger_index')
 
 /*
@@ -48,7 +48,8 @@ workflow NFDATAOMICS_SPATIALOMICS {
     ch_fasta              = params.fasta             ? Channel.value(file(params.fasta, checkIfExists: true))             : Channel.empty()
     ch_gtf                = params.gtf               ? Channel.value(file(params.gtf, checkIfExists: true))               : Channel.empty()
     ch_gff                = params.gff               ? Channel.value(file(params.gff, checkIfExists: true))               : Channel.empty()
-    ch_spaceranger_index  = params.spaceranger_index ? Channel.value(file(params.spaceranger_index, checkIfExists: true)) : Channel.empty()
+    ch_spaceranger_index  = params.spaceranger_index ? file(params.spaceranger_index, checkIfExists: true) : Channel.empty()
+    ch_probeset           = params.probeset          ? Channel.value(file(params.probeset, checkIfExists: true))          : Channel.empty()
 
     //
     // WORKFLOW: Run pipeline
@@ -58,7 +59,8 @@ workflow NFDATAOMICS_SPATIALOMICS {
         ch_fasta,
         ch_gtf,
         ch_gff,
-        ch_spaceranger_index
+        ch_spaceranger_index,
+        ch_probeset
     )
     emit:
     multiqc_report = SPATIALOMICS.out.multiqc_report // channel: /path/to/multiqc_report.html

modules.json

@@ -35,6 +35,17 @@
                         "git_sha": "e10b76ca0c66213581bec2833e30d31f239dec0b",
                         "installed_by": ["modules"]
                     },
+                    "spaceranger/count": {
+                        "branch": "master",
+                        "git_sha": "41dfa3f7c0ffabb96a6a813fe321c6d1cc5b6e46",
+                        "installed_by": ["modules"],
+                        "patch": "modules/nf-core/spaceranger/count/spaceranger-count.diff"
+                    },
+                    "spaceranger/mkgtf": {
+                        "branch": "master",
+                        "git_sha": "41dfa3f7c0ffabb96a6a813fe321c6d1cc5b6e46",
+                        "installed_by": ["modules"]
+                    },
                     "spaceranger/mkref": {
                         "branch": "master",
                         "git_sha": "41dfa3f7c0ffabb96a6a813fe321c6d1cc5b6e46",