Merge remote-tracking branch 'upstream/master' into main

modules/deepvariant/main.nf

@@ -4,10 +4,10 @@ process DEEPVARIANT {
 
 
     if (params.enable_conda) {
-        exit 1, "Conda environments cannot be used when using the DeepVariant tool. Please use docker or singularity containers."
+        exit 1, "Conda environments cannot be used with DeepVariant at the moment. Please use Docker or Singularity containers."
     }
 
-    container 'google/deepvariant:1.3.0'
+    container "google/deepvariant:1.3.0"
 
     input:
     tuple val(meta), path(input), path(index), path(intervals)

modules/shasta/main.nf

@@ -0,0 +1,49 @@
+process SHASTA {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda (params.enable_conda ? "bioconda::shasta=0.8.0" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/shasta:0.8.0--h7d875b9_0':
+        'quay.io/biocontainers/shasta:0.8.0--h7d875b9_0' }"
+
+    input:
+    tuple val(meta), path(reads)
+
+    output:
+    tuple val(meta), path("*_Assembly.fasta.gz"), emit: assembly
+    tuple val(meta), path("*_Assembly.gfa.gz")  , emit: gfa
+    tuple val(meta), path("ShastaRun/")                 , emit: results
+    path "versions.yml"                                 , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args   = task.ext.args   ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def model  = "${meta.model}" ?: 'Nanopore-Oct2021'
+    """
+    # shasta requires uncompressed
+    zcat -f $reads > reads.fq
+
+    # run shasta
+    shasta \\
+        --input reads.fq \\
+        --config $model \\
+        $args \\
+        --threads $task.cpus
+
+    # compress results
+    gzip -c ShastaRun/Assembly.fasta > ${prefix}_Assembly.fasta.gz
+    gzip -c ShastaRun/Assembly.gfa   > ${prefix}_Assembly.gfa.gz
+
+    # cleanup temp files
+    rm reads.fq
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        shasta: \$(shasta --version | head -n 1 | cut -f 3 -d " ")
+    END_VERSIONS
+    """
+}

modules/shasta/meta.yml

@@ -0,0 +1,52 @@
+name: shasta
+description: The goal of the Shasta long read assembler is to rapidly produce accurate assembled sequence using DNA reads generated by Oxford Nanopore flow cells as input. Please note Assembler is design to focus on speed, so assembly may be considered somewhat non-deterministic as final assembly may vary across executions. See https://github.com/chanzuckerberg/shasta/issues/296.
+keywords:
+  - nanopore
+  - de-novo
+  - assembly
+  - longread
+tools:
+  - shasta:
+      description: Rapidly produce accurate assembled sequence using as input DNA reads generated by Oxford Nanopore flow cells.
+      homepage: https://chanzuckerberg.github.io/shasta/index.html
+      documentation: https://chanzuckerberg.github.io/shasta/index.html
+      tool_dev_url: https://github.com/chanzuckerberg/shasta
+      doi: "https://doi.org/10.1038/s41587-020-0503-6"
+      licence: ["MIT"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - reads:
+      type: file
+      description: Input file in FASTQ format.
+      pattern: "*.{fastq,fastq.gz,fq,fq.gz}"
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - assembly:
+      type: file
+      description: Assembled FASTA file
+      pattern: "${prefix}_Assembly.fasta.gz"
+  - gfa:
+      type: file
+      description: Repeat graph
+      pattern: "${prefix}_Assembly.gfa.gz"
+  - results:
+      type: dir
+      description: Resulting assembly directory
+      pattern: "ShastaRun"
+
+authors:
+  - "@fmalmeida"

tests/config/pytest_modules.yml

@@ -2287,6 +2287,10 @@ sexdeterrmine:
   - modules/sexdeterrmine/**
   - tests/modules/sexdeterrmine/**
 
+shasta:
+  - modules/shasta/**
+  - tests/modules/shasta/**
+
 shasum:
   - modules/shasum/**
   - tests/modules/shasum/**

tests/modules/shasta/main.nf

@@ -0,0 +1,15 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { SHASTA } from '../../../modules/shasta/main.nf'
+
+workflow test_shasta {
+
+    input = [ 
+        [ id:'test', model:'Nanopore-Dec2019' ], // meta map
+        [ file(params.test_data['candidatus_portiera_aleyrodidarum']['nanopore']['test_fastq_gz']) ],
+    ]
+
+    SHASTA ( input )
+}

tests/modules/shasta/nextflow.config

@@ -0,0 +1,7 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+
+    ext.args   = '--Reads.minReadLength 100'
+
+}

tests/modules/shasta/test.yml

@@ -0,0 +1,35 @@
+- name: shasta test_shasta
+  command: nextflow run ./tests/modules/shasta -entry test_shasta -c ./tests/config/nextflow.config  -c ./tests/modules/shasta/nextflow.config
+  tags:
+    - shasta
+  files:
+    - path: output/shasta/ShastaRun/Assembly-BothStrands-NoSequence.gfa
+    - path: output/shasta/ShastaRun/Assembly-BothStrands.gfa
+    - path: output/shasta/ShastaRun/Assembly.fasta
+    - path: output/shasta/ShastaRun/Assembly.gfa
+    - path: output/shasta/ShastaRun/AssemblySummary.csv
+    - path: output/shasta/ShastaRun/AssemblySummary.html
+    - path: output/shasta/ShastaRun/AssemblySummary.json
+    - path: output/shasta/ShastaRun/Binned-ReadLengthHistogram.csv
+      md5sum: f7997650e6ff2b060c583c6fa796628e
+    - path: output/shasta/ShastaRun/DisjointSetsHistogram.csv
+      md5sum: eecb9fae34b719dad64952257e2398a6
+    - path: output/shasta/ShastaRun/LowHashBucketHistogram.csv
+      md5sum: e5eb4a0d28ade913de6584928f32b7aa
+    - path: output/shasta/ShastaRun/MarkerGraphEdgeCoverageHistogram.csv
+      md5sum: 37ecac444067fbf4e0d96dc2705a9ada
+    - path: output/shasta/ShastaRun/MarkerGraphVertexCoverageHistogram.csv
+      md5sum: ccb236ab97ab1b9964b4856aa47d3c94
+    - path: output/shasta/ShastaRun/ReadGraphComponents.csv
+    - path: output/shasta/ShastaRun/ReadLengthHistogram.csv
+      md5sum: ccb25c4850ba2682fd1d452d6596d7bc
+    - path: output/shasta/ShastaRun/ReadLowHashStatistics.csv
+      md5sum: 2431d47341b239cb6edf8e3a7a76021f
+    - path: output/shasta/ShastaRun/ReadSummary.csv
+      md5sum: 8e0bffccf7afae54a937266949d118af
+    - path: output/shasta/ShastaRun/index.html
+      md5sum: 2d067ccc873b94a73e7300fe597db135
+    - path: output/shasta/ShastaRun/shasta.conf
+      md5sum: cba4aebac8a74eb5868f4a20f04bc618
+    - path: output/shasta/test_Assembly.fasta.gz
+    - path: output/shasta/test_Assembly.gfa.gz