From 97fbd3264485b2c2fe719bc3c1bcd9d4476eda86 Mon Sep 17 00:00:00 2001
From: Phil Ewels <phil.ewels@scilifelab.se>
Date: Thu, 5 Dec 2019 15:35:51 +0100
Subject: [PATCH] Added samtools sort + index

Added container to FastQC
---
 .github/workflows/test-processes.yml |  8 ++++++
 tools/fastqc/main.nf                 | 38 +++++++++++++++-------------
 tools/fastqc/meta.yml                |  2 +-
 tools/samtools/index/main.nf         | 22 ++++++++++++++++
 tools/samtools/index/meta.yml        | 27 ++++++++++++++++++++
 tools/samtools/index/test/main.nf    |  1 +
 tools/samtools/sort/main.nf          | 18 +++++++++++++
 tools/samtools/sort/meta.yml         | 27 ++++++++++++++++++++
 tools/samtools/sort/test/main.nf     |  1 +
 9 files changed, 125 insertions(+), 19 deletions(-)
 create mode 100644 tools/samtools/index/main.nf
 create mode 100644 tools/samtools/index/meta.yml
 create mode 100644 tools/samtools/index/test/main.nf
 create mode 100644 tools/samtools/sort/main.nf
 create mode 100644 tools/samtools/sort/meta.yml
 create mode 100644 tools/samtools/sort/test/main.nf

diff --git a/.github/workflows/test-processes.yml b/.github/workflows/test-processes.yml
index b9c56955..2fb14e95 100644
--- a/.github/workflows/test-processes.yml
+++ b/.github/workflows/test-processes.yml
@@ -16,3 +16,11 @@ jobs:
     - name: FastQC
       path: tools/fastqc/*
       run: nextflow run ./tools/fastqc/test/
+
+    - name: samtools sort
+      path: tools/samtoosl/sort*
+      run: nextflow run ./tools/samtools/sort/test/
+
+    - name: samtools index
+      path: tools/samtoosl/index*
+      run: nextflow run ./tools/samtools/index/test/
diff --git a/tools/fastqc/main.nf b/tools/fastqc/main.nf
index deb570ff..6bc53a9c 100644
--- a/tools/fastqc/main.nf
+++ b/tools/fastqc/main.nf
@@ -1,18 +1,20 @@
-process fastqc {
-    tag "$sample_id"
-    publishDir "${params.outdir}/fastqc", mode: 'copy',
-        saveAs: {filename -> filename.indexOf(".zip") > 0 ? "zips/$filename" : "$filename"}
-
-    input:
-    set val(sample_id), file(reads)
-
-    output:
-    file "*_fastqc.{zip,html}"
-
-    script:
-    """
-    fastqc -q $reads
-
-    fastqc --version &> fastqc.version.txt
-    """
-}
+process fastqc {
+    tag "$sample_id"
+    publishDir "${params.outdir}/fastqc", mode: 'copy',
+        saveAs: {filename -> filename.indexOf(".zip") > 0 ? "zips/$filename" : "$filename"}
+
+        container: 'quay.io/biocontainers/fastqc:0.11.8--2'
+
+    input:
+    set val(sample_id), file(reads)
+
+    output:
+    file "*_fastqc.{zip,html}"
+
+    script:
+    """
+    fastqc -q $reads
+
+    fastqc --version &> fastqc.version.txt
+    """
+}
diff --git a/tools/fastqc/meta.yml b/tools/fastqc/meta.yml
index ddd0ae7f..0d5afc23 100644
--- a/tools/fastqc/meta.yml
+++ b/tools/fastqc/meta.yml
@@ -12,7 +12,7 @@ tools:
             across your reads, the per base sequence content (%A/C/G/T).
             You get information about adapter contamination and other
             overrepresented sequences.
-        homepage: hhttps://www.bioinformatics.babraham.ac.uk/projects/fastqc/
+        homepage: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
         documentation: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/Help/
 input:
     -
diff --git a/tools/samtools/index/main.nf b/tools/samtools/index/main.nf
new file mode 100644
index 00000000..8d29f824
--- /dev/null
+++ b/tools/samtools/index/main.nf
@@ -0,0 +1,22 @@
+process samtools_index {
+    tag "${bam.baseName}"
+
+    container: 'quay.io/biocontainers/samtools:1.9--h10a08f8_12'
+
+    input:
+    set file(bam)
+
+    output:
+    file "*.sorted.bam"
+
+    script:
+    def suff_mem = ("${(task.memory.toBytes() - 6000000000) / task.cpus}" > 2000000000) ? 'true' : 'false'
+    def avail_mem = (task.memory && suff_mem) ? "-m" + "${(task.memory.toBytes() - 6000000000) / task.cpus}" : ''
+    """
+    samtools sort $bam \\
+        -@ ${task.cpus} ${avail_mem} \\
+        -o ${bam.baseName}.sorted.bam
+
+    samtools --version &> v_samtools.txt
+    """
+}
diff --git a/tools/samtools/index/meta.yml b/tools/samtools/index/meta.yml
new file mode 100644
index 00000000..3cd7a402
--- /dev/null
+++ b/tools/samtools/index/meta.yml
@@ -0,0 +1,27 @@
+name: samtools sort
+description: Sort a BAM or CRAM file
+keywords:
+    - sort
+tools:
+    - samtools:
+        description: |
+            SAMtools is a set of utilities for interacting with and post-processing
+            short DNA sequence read alignments in the SAM, BAM and CRAM formats, written by Heng Li.
+            These files are generated as output by short read aligners like BWA.
+        homepage: http://www.htslib.org/
+        documentation: hhttp://www.htslib.org/doc/samtools.html
+        doi: 10.1093/bioinformatics/btp352
+input:
+    -
+        - input:
+            type: file
+            description: Input BAM or CRAM file
+            pattern: *.{bam,cram}
+output:
+    -
+        - index:
+            type: file
+            description: BAM or CRAM index file
+            pattern: *.{bai}
+authors:
+    - @ewels
diff --git a/tools/samtools/index/test/main.nf b/tools/samtools/index/test/main.nf
new file mode 100644
index 00000000..2e9ae0f5
--- /dev/null
+++ b/tools/samtools/index/test/main.nf
@@ -0,0 +1 @@
+/* A mini pipeline to test samtools index here */
diff --git a/tools/samtools/sort/main.nf b/tools/samtools/sort/main.nf
new file mode 100644
index 00000000..696c8872
--- /dev/null
+++ b/tools/samtools/sort/main.nf
@@ -0,0 +1,18 @@
+process samtools_index {
+    tag "${bam.baseName}"
+
+    container: 'quay.io/biocontainers/samtools:1.9--h10a08f8_12'
+
+    input:
+    set file(bam)
+
+    output:
+    file "*.bam.bai"
+
+    script:
+    """
+    samtools index $bam
+
+    samtools --version &> v_samtools.txt
+    """
+}
diff --git a/tools/samtools/sort/meta.yml b/tools/samtools/sort/meta.yml
new file mode 100644
index 00000000..60f9e96b
--- /dev/null
+++ b/tools/samtools/sort/meta.yml
@@ -0,0 +1,27 @@
+name: samtools sort
+description: Sort a BAM or CRAM file
+keywords:
+    - sort
+tools:
+    - samtools:
+        description: |
+            SAMtools is a set of utilities for interacting with and post-processing
+            short DNA sequence read alignments in the SAM, BAM and CRAM formats, written by Heng Li.
+            These files are generated as output by short read aligners like BWA.
+        homepage: http://www.htslib.org/
+        documentation: hhttp://www.htslib.org/doc/samtools.html
+        doi: 10.1093/bioinformatics/btp352
+input:
+    -
+        - input:
+            type: file
+            description: Input BAM or CRAM file
+            pattern: *.{bam,cram}
+output:
+    -
+        - sorted_file:
+            type: file
+            description: Sorted BAM or CRAM file
+            pattern: *.{bam,cram}
+authors:
+    - @ewels
diff --git a/tools/samtools/sort/test/main.nf b/tools/samtools/sort/test/main.nf
new file mode 100644
index 00000000..6e0055e2
--- /dev/null
+++ b/tools/samtools/sort/test/main.nf
@@ -0,0 +1 @@
+/* A mini pipeline to test samtools sort here */