diamond blastx subworkflow

conf/modules.config

@@ -37,12 +37,20 @@ process {
         ext.args = "--evalue 1.0e-25  --max-target-seqs 10 --max-hsps 1"
     }
 
+    withName: "DIAMOND_BLASTX" {
+        ext.args = "--evalue 1.0e-25  --max-target-seqs 10 --max-hsps 1"
+    }
+
     withName: "BLOBTOOLKIT_WINDOWSTATS" {
         ext.args = "--window 0.1 --window 0.01 --window 1 --window 100000 --window 1000000"
     }
 
-    withName: "BLOBTOOLKIT_BLOBDIR" {
+    withName: "BLOBTOOLKIT_CREATEBLOBDIR" {
         ext.args = "--evalue 1.0e-25 --hit-count 10"
+    }
+
+    withName: "BLOBTOOLKIT_CREATEBLOBDIR" {
+        ext.args = "--evalue 1.0e-25 --hit-count 10 --update-plot"
         publishDir = [
             path: { "${params.outdir}/" },
             mode: params.publish_dir_mode,
@@ -66,6 +74,14 @@ process {
         ]
     }
 
+    withName: "BLOBTOOLKIT_CHUNK" {
+        ext.args = "--chunk 100000 --overlap 0 --max-chunks 10 --min-length 1000"
+    }
+
+    withName: "BLOBTOOLKIT_UNCHUNK" {
+        ext.args = "--count 10"
+    }
+
     withName: "CUSTOM_DUMPSOFTWAREVERSIONS" {
         publishDir = [
             path: { "${params.outdir}/blobtoolkit_info" },

conf/test.config

@@ -30,7 +30,8 @@ params {
     taxon     = "Meles meles"
 
     // Databases
-    taxdump   = "/lustre/scratch123/tol/teams/grit/geval_pipeline/btk_databases/taxdump"
-    busco     = "/lustre/scratch123/tol/resources/nextflow/busco_2021_06_reduced/"
-    uniprot   = "${projectDir}/assets/test/mCerEla1.1.buscogenes.dmnd"
+    taxdump = "/lustre/scratch123/tol/teams/grit/geval_pipeline/btk_databases/taxdump"
+    busco   = "/lustre/scratch123/tol/resources/nextflow/busco_2021_06_reduced/"
+    blastp  = "${projectDir}/assets/test/mMelMel3.1.buscogenes.dmnd"
+    blastx  = "${projectDir}/assets/test/mMelMel3.1.buscoregions.dmnd"
 }

conf/test_full.config

@@ -27,7 +27,9 @@ params {
     taxon     = "Laetiporus sulphureus"
 
     // Databases
-    taxdump   = "/lustre/scratch123/tol/teams/grit/geval_pipeline/btk_databases/taxdump"
-    busco     = "/lustre/scratch123/tol/resources/busco/v5/"
-    uniprot   = "${projectDir}/assets/test_full/gfLaeSulp1.1.buscogenes.dmnd"
+    taxdump = "/lustre/scratch123/tol/teams/grit/geval_pipeline/btk_databases/taxdump"
+    busco   = "/lustre/scratch123/tol/resources/busco/v5/"
+    blastp  = "${projectDir}/assets/test_full/gfLaeSulp1.1.buscogenes.dmnd"
+    blastx  = "${projectDir}/assets/test_full/gfLaeSulp1.1.buscoregions.dmnd"
+
 }

modules.json

@@ -21,6 +21,11 @@
                         "git_sha": "911696ea0b62df80e900ef244d7867d177971f73",
                         "installed_by": ["modules"]
                     },
+                    "diamond/blastx": {
+                        "branch": "master",
+                        "git_sha": "911696ea0b62df80e900ef244d7867d177971f73",
+                        "installed_by": ["modules"]
+                    },
                     "fastawindows": {
                         "branch": "master",
                         "git_sha": "911696ea0b62df80e900ef244d7867d177971f73",

modules/local/blobtoolkit/chunk.nf

@@ -0,0 +1,36 @@
+process BLOBTOOLKIT_CHUNK {
+    tag "$meta.id"
+    label 'process_single'
+
+    if (workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1) {
+        exit 1, "BLOBTOOLKIT_CHUNK module does not support Conda. Please use Docker / Singularity / Podman instead."
+    }
+    container "genomehubs/blobtoolkit:4.1.5"
+
+    input:
+    tuple val(meta) , path(fasta)
+    tuple val(meta2), path(busco_table)
+
+    output:
+    tuple val(meta), path("*.chunks.fasta"), emit: chunks
+    path "versions.yml"                    , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    btk pipeline chunk-fasta \\
+        --in ${fasta} \\
+        --busco ${busco_table} \\
+        --out ${prefix}.chunks.fasta \\
+        $args
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        blobtoolkit: \$(btk --version | cut -d' ' -f2 | sed 's/v//')
+    END_VERSIONS
+    """
+}

modules/local/blobtoolkit/blobdir.nf → modules/local/blobtoolkit/createblobdir.nf

@@ -1,4 +1,4 @@
-process BLOBTOOLKIT_BLOBDIR {
+process BLOBTOOLKIT_CREATEBLOBDIR {
     tag "$meta.id"
     label 'process_medium'
 

modules/local/blobtoolkit/unchunk.nf

@@ -0,0 +1,34 @@
+process BLOBTOOLKIT_UNCHUNK {
+    tag "$meta.id"
+    label 'process_single'
+
+    if (workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1) {
+        exit 1, "BLOBTOOLKIT_UNCHUNK module does not support Conda. Please use Docker / Singularity / Podman instead."
+    }
+    container "genomehubs/blobtoolkit:4.1.5"
+
+    input:
+    tuple val(meta), path(blast_table)
+
+    output:
+    tuple val(meta), path("*.blastx.out"), emit: blastx
+    path "versions.yml"                  , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    btk pipeline unchunk-blast \\
+        --in ${blast_table} \\
+        --out ${prefix}.blastx.out \\
+        $args
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        blobtoolkit: \$(btk --version | cut -d' ' -f2 | sed 's/v//')
+    END_VERSIONS
+    """
+}

modules/local/blobtoolkit/updateblobdir.nf

@@ -0,0 +1,42 @@
+process BLOBTOOLKIT_UPDATEBLOBDIR {
+    tag "$meta.id"
+    label 'process_medium'
+
+    if (workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1) {
+        exit 1, "BLOBTOOLKIT_BLOBDIR module does not support Conda. Please use Docker / Singularity / Podman instead."
+    }
+    container "genomehubs/blobtoolkit:4.1.5"
+
+    input:
+    tuple val(meta), path(input)
+    tuple val(meta1), path(blastx)
+    path(taxdump)
+
+    output:
+    tuple val(meta), path(prefix), emit: blobdir
+    path "versions.yml"          , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    prefix = task.ext.prefix ?: "${meta.id}"
+    def hits_blastx = blastx ? "--hits ${blastx}" : ""
+    """
+    blobtools replace \\
+        --taxdump ${taxdump} \\
+        --taxrule bestdistorder=buscoregions \\
+        ${hits_blastx} \\
+        --bedtsvdir windowstats \\
+        --meta ${yaml} \\
+        --threads ${task.cpus} \\
+        $args \\
+        ${input}
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        blobtoolkit: \$(btk --version | cut -d' ' -f2 | sed 's/v//')
+    END_VERSIONS
+    """
+}

nextflow.config

@@ -23,9 +23,13 @@ params {
     // Databases and related options
     taxdump                    = null
     busco                      = null
-    uniprot                    = null
+    blastp                     = null
+    blastx                     = null
     blastp_outext              = 'txt'
     blastp_cols                = 'qseqid staxids bitscore qseqid sseqid pident length mismatch gapopen qstart qend sstart send evalue bitscore'
+    blastx_outext              = 'txt'
+    blastx_cols                = 'qseqid staxids bitscore qseqid sseqid pident length mismatch gapopen qstart qend sstart send evalue bitscore'
+
 
     // MultiQC options
     multiqc_config             = null