testing bam input and mapping to transcriptome

Author: number-25

assets/test_data/KCMF1_hg38_transcript.fa.fai

@@ -0,0 +1 @@
+XM_047445126.1	7618	115	70	71

conf/test.config

@@ -32,6 +32,7 @@ params {
     genome_fasta_sizes              = "assets/test_data/KCMF1_hg38.fa.sizes"
     genome_minimap2_index           = "assets/test_data/KCMF1_hg38_k14.mmi"
     transcriptome_fasta             = "assets/test_data/KCMF1_hg38_transcript.fa"
+    transcriptome_fasta_index       = "assets/test_data/KCMF1_hg38_transcript.fa.fai"
     transcriptome_minimap2_index    = "assets/test_data/KCMF1_hg38_transcript_k14.mmi"
     annotation_gtf                  = "assets/test_data/KCMF1_hg38_transcript.gtf"
     skip_prepare_reference          = true

conf/test_full.config

@@ -29,6 +29,7 @@ params {
     genome_fasta_sizes              = null
     genome_minimap2_index           = null
     transcriptome_fasta             = "assets/fulltest_data/Homo_sapiens.GRCh38.112_chr_transcripts.fa.gz"
+    transcriptome_fasta_index       = null
     transcriptome_minimap2_index    = null
     annotation_gtf                  = "assets/fulltest_data/Homo_sapiens.GRCh38.112_chr.gtf.gz"
     skip_prepare_reference          = false

modules/local/samtools/fasta/main.nf

@@ -0,0 +1,48 @@
+process SAMTOOLS_FASTA {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/samtools:1.20--h50ea8bc_0' :
+        'biocontainers/samtools:1.20--h50ea8bc_0' }"
+
+    input:
+    tuple val(meta), path(bam)
+
+    output:
+    tuple val(meta), path("*.fasta"),       emit: fasta
+    path "versions.yml",                    emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args      = task.ext.args ?: ''
+    def prefix    = task.ext.prefix ?: "${meta.id}_${meta.replicate}_unmapped"
+
+    """
+    samtools fasta \\
+        -f 4 \\
+        -@ $task.cpus \\
+        ${bam} \\
+        > ${prefix}.fasta \\
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        samtools: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//')
+    END_VERSIONS
+    """
+
+    stub:
+    def args      = task.ext.args ?: ''
+    def prefix    = task.ext.prefix ?: "${meta.id}_${meta.replicate}_unmapped"
+
+    """
+    touch ${prefix}.fasta
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        samtools: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//')
+    END_VERSIONS
+    """
+}

modules/local/samtools/fasta/nextflow.config

@@ -0,0 +1,8 @@
+process {
+    withName: 'SAMTOOLS_FASTA' {
+        publishDir = [
+        path: { "${params.outdir}/mapping/unmapped_reads/${meta.id}_${meta.replicate}" },
+        //mode: params.publish_dir_mode
+        ]
+    }
+}

modules/local/sylph/profile/main.nf

@@ -23,13 +23,12 @@ process SYLPH_PROFILE {
     """
     sylph profile \\
         $sylph_database \\
-        $sample \\
-        $task.cpus \\
+        -r $sample \\
         -o ${prefix}_sylph.tsv
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
-        sylph_profile: \$(sylph --version | cut -d" " -f2')
+        sylph_profile: \$(sylph --version | cut -d" " -f2)
     END_VERSIONS
     """
 
@@ -40,7 +39,7 @@ process SYLPH_PROFILE {
     touch ${prefix}_sylph.tsv
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
-        sylph_profile: \$(sylph --version | cut -d" " -f2')
+        sylph_profile: \$(sylph --version | cut -d" " -f2)
     END_VERSIONS
     """
 }

modules/local/sylph/tax/main.nf

@@ -21,6 +21,8 @@ process SYLPH_TAX {
     def prefix = task.ext.prefix ?: "${meta.id}_${meta.replicate}"
 
     """
+    sylph-tax download --download-to .
+
     sylph-tax taxprof \\
         $sylph_profile \\
         -t $database_name \\

nextflow.config

@@ -24,6 +24,7 @@ params {
     skip_prepare_reference          = false // even if paths are hardcoded, files need to be unzipped and so on
     genome_minimap2_index           = null
     transcriptome_fasta             = null // e.g. 'data/gencode.v46.transcripts.fa.gz'
+    transcriptome_fasta_index       = null
     transcriptome_minimap2_index    = null
     annotation_gtf                  = null // e.g. 'data/Homo_sapiens.GRCh38.112.gtf.gz'
 

nextflow_schema.json

@@ -35,6 +35,10 @@
                     "type": "string",
                     "description": "Path to transcriptome fasta sequence."
                 },
+                "transcriptome_fasta_index": {
+                    "type": "string",
+                    "description": "Path to transcriptome fasta index."
+                },
                 "transcriptome_minimap2_index": {
                     "type": "string",
                     "description": "Path to transcriptome minimap2 index."