From 504028346bc4b7c37bab0f40d437e4fa7b45904f Mon Sep 17 00:00:00 2001
From: Priyanka Surana <86618929+priyanka-surana@users.noreply.github.com>
Date: Thu, 29 Sep 2022 08:10:44 +0100
Subject: [PATCH] samtools fasta module (#2128)

---
 modules/samtools/fasta/main.nf               | 37 ++++++++++++++++
 modules/samtools/fasta/meta.yml              | 46 ++++++++++++++++++++
 tests/config/pytest_modules.yml              |  4 ++
 tests/modules/samtools/fasta/main.nf         | 15 +++++++
 tests/modules/samtools/fasta/nextflow.config |  5 +++
 tests/modules/samtools/fasta/test.yml        | 10 +++++
 6 files changed, 117 insertions(+)
 create mode 100644 modules/samtools/fasta/main.nf
 create mode 100644 modules/samtools/fasta/meta.yml
 create mode 100644 tests/modules/samtools/fasta/main.nf
 create mode 100644 tests/modules/samtools/fasta/nextflow.config
 create mode 100644 tests/modules/samtools/fasta/test.yml

diff --git a/modules/samtools/fasta/main.nf b/modules/samtools/fasta/main.nf
new file mode 100644
index 00000000..130c2a5a
--- /dev/null
+++ b/modules/samtools/fasta/main.nf
@@ -0,0 +1,37 @@
+process SAMTOOLS_FASTA {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda (params.enable_conda ? "bioconda::samtools=1.15.1" : null)
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/samtools:1.15.1--h1170115_0':
+        'quay.io/biocontainers/samtools:1.15.1--h1170115_0' }"
+
+    input:
+    tuple val(meta), path(input)
+
+    output:
+    tuple val(meta), path("*.fasta.gz"), 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}"
+    def endedness = meta.single_end ? "-0 ${prefix}.fasta.gz" : "-1 ${prefix}_1.fasta.gz -2 ${prefix}_2.fasta.gz"
+    """
+    samtools \\
+        fasta \\
+        $args \\
+        --threads ${task.cpus-1} \\
+        $endedness \\
+        $input
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        samtools: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//' ))
+    END_VERSIONS
+    """
+}
diff --git a/modules/samtools/fasta/meta.yml b/modules/samtools/fasta/meta.yml
new file mode 100644
index 00000000..4c662269
--- /dev/null
+++ b/modules/samtools/fasta/meta.yml
@@ -0,0 +1,46 @@
+name: "samtools_fasta"
+description: Converts a SAM/BAM/CRAM file to FASTA
+keywords:
+  - bam
+  - sam
+  - cram
+  - fasta
+tools:
+  - "samtools":
+      description: "Tools for dealing with SAM, BAM and CRAM files"
+      homepage: "http://www.htslib.org"
+      documentation: "https://www.htslib.org/doc/samtools-fasta.html"
+      tool_dev_url: "https://github.com/samtools/samtools"
+      doi: "10.1093/bioinformatics/btp352"
+      licence: "['MIT']"
+
+input:
+  # Only when we have meta
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - input:
+      type: file
+      description: BAM/CRAM/SAM file
+      pattern: "*.{bam,cram,sam}"
+
+output:
+  #Only when we have meta
+  - 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"
+  - fasta:
+      type: file
+      description: compressed FASTA file
+      pattern: "*.fasta.gz"
+
+authors:
+  - "@priyanka-surana"
diff --git a/tests/config/pytest_modules.yml b/tests/config/pytest_modules.yml
index 16156bbf..cae10d10 100644
--- a/tests/config/pytest_modules.yml
+++ b/tests/config/pytest_modules.yml
@@ -2103,6 +2103,10 @@ samtools/faidx:
   - modules/samtools/faidx/**
   - tests/modules/samtools/faidx/**
 
+samtools/fasta:
+  - modules/samtools/fasta/**
+  - tests/modules/samtools/fasta/**
+
 samtools/fastq:
   - modules/samtools/fastq/**
   - tests/modules/samtools/fastq/**
diff --git a/tests/modules/samtools/fasta/main.nf b/tests/modules/samtools/fasta/main.nf
new file mode 100644
index 00000000..c0ef91f4
--- /dev/null
+++ b/tests/modules/samtools/fasta/main.nf
@@ -0,0 +1,15 @@
+#!/usr/bin/env nextflow
+
+nextflow.enable.dsl = 2
+
+include { SAMTOOLS_FASTA } from '../../../../modules/samtools/fasta/main.nf'
+
+workflow test_samtools_fasta {
+    
+    input = [
+        [ id:'test', single_end:false ], // meta map
+        file(params.test_data['sarscov2']['illumina']['test_paired_end_bam'], checkIfExists: true)
+    ]
+
+    SAMTOOLS_FASTA ( input )
+}
diff --git a/tests/modules/samtools/fasta/nextflow.config b/tests/modules/samtools/fasta/nextflow.config
new file mode 100644
index 00000000..50f50a7a
--- /dev/null
+++ b/tests/modules/samtools/fasta/nextflow.config
@@ -0,0 +1,5 @@
+process {
+
+    publishDir = { "${params.outdir}/${task.process.tokenize(':')[-1].tokenize('_')[0].toLowerCase()}" }
+    
+}
\ No newline at end of file
diff --git a/tests/modules/samtools/fasta/test.yml b/tests/modules/samtools/fasta/test.yml
new file mode 100644
index 00000000..6c992caf
--- /dev/null
+++ b/tests/modules/samtools/fasta/test.yml
@@ -0,0 +1,10 @@
+- name: samtools fasta test_samtools_fasta
+  command: nextflow run ./tests/modules/samtools/fasta -entry test_samtools_fasta -c ./tests/config/nextflow.config  -c ./tests/modules/samtools/fasta/nextflow.config
+  tags:
+    - samtools
+    - samtools/fasta
+  files:
+    - path: output/samtools/test_1.fasta.gz
+      md5sum: 3abd682290bc7c75f1ce2b80db995237
+    - path: output/samtools/test_2.fasta.gz
+      md5sum: 6ada09ce66f68b8732985e14aac1bf1f