Merge branch 'main' into samtools_sort

CHANGELOG.md

@@ -37,7 +37,8 @@
 
 * `samtools`:
     - `samtools/samtools_flagstat`: Counts the number of alignments in SAM/BAM/CRAM files for each FLAG type (PR #31).
-    - `samtools/samtools_idxstats`: Reports alignment summary statistics for a SAM/BAM/CRAM file (PR #32).    
+    - `samtools/samtools_idxstats`: Reports alignment summary statistics for a SAM/BAM/CRAM file (PR #32).
+    - `samtools/samtools_index`: Index SAM/BAM/CRAM files (PR #35).
     - `samtools/samtools_sort`: Sort SAM/BAM/CRAM files (PR #36).
 
 ## MAJOR CHANGES

src/gffread/script.sh

@@ -51,7 +51,7 @@
 [[ "$par_cluster_only" == "false" ]] && unset par_cluster_only
 
 
-gffread \
+$(which gffread) \
     "$par_input" \
     ${par_chr_mapping:+-m "$par_chr_mapping"} \
     ${par_seq_info:+-s "$par_seq_info"} \

src/samtools/samtools_index/config.vsh.yaml

@@ -0,0 +1,67 @@
+name: samtools_index
+namespace: samtools
+description: Index SAM/BAM/CRAM files.
+keywords: [index, bam, sam, cram]
+links:
+  homepage: https://www.htslib.org/
+  documentation: https://www.htslib.org/doc/samtools-index.html
+  repository: https://github.com/samtools/samtools
+references: 
+  doi: [10.1093/bioinformatics/btp352, 10.1093/gigascience/giab008]
+license: MIT/Expat
+
+argument_groups:
+  - name: Inputs
+    arguments:
+      - name: --input
+        type: file
+        description: Input file name
+        required: true
+        must_exist: true
+  - name: Outputs
+    arguments:
+      - name: --output
+        alternatives: -o
+        type: file
+        description: Output file name
+        required: true
+        direction: output
+        example: out.bam.bai
+  - name: Options
+    arguments:
+      - name: --bai
+        alternatives: -b
+        type: boolean_true
+        description: Generate BAM index
+      - name: --csi
+        alternatives: -c
+        type: boolean_true
+        description: |
+          Create a CSI index for BAM files instead of the traditional BAI 
+          index. This will be required for genomes with larger chromosome 
+          sizes.
+      - name: --min_shift
+        alternatives: -m
+        type: integer
+        description: |
+          Create a CSI index, with a minimum interval size of 2^INT.
+
+resources:
+  - type: bash_script
+    path: script.sh
+test_resources:
+  - type: bash_script
+    path: test.sh
+  - type: file
+    path: test_data
+engines:
+  - type: docker
+    image: quay.io/biocontainers/samtools:1.19.2--h50ea8bc_1
+    setup:
+      - type: docker
+        run: |
+          samtools --version 2>&1 | grep -E '^(samtools|Using htslib)' | \
+          sed 's#Using ##;s# \([0-9\.]*\)$#: \1#' > /var/software_versions.txt
+runners:
+- type: executable
+- type: nextflow

src/samtools/samtools_index/help.txt

@@ -0,0 +1,13 @@
+```
+samtools index
+```
+
+Usage: samtools index -M [-bc] [-m INT] <in1.bam> <in2.bam>...
+   or: samtools index [-bc] [-m INT] <in.bam> [out.index]
+Options:
+  -b, --bai            Generate BAI-format index for BAM files [default]
+  -c, --csi            Generate CSI-format index for BAM files
+  -m, --min-shift INT  Set minimum interval size for CSI indices to 2^INT [14]
+  -M                   Interpret all filename arguments as files to be indexed
+  -o, --output FILE    Write index to FILE [alternative to <out.index> in args]
+  -@, --threads INT    Sets the number of threads [none]

src/samtools/samtools_index/script.sh

@@ -0,0 +1,18 @@
+#!/bin/bash
+
+## VIASH START
+## VIASH END
+
+set -e
+[[ "$par_multiple" == "false" ]] && unset par_multiple
+[[ "$par_bai" == "false" ]] && unset par_bai
+[[ "$par_csi" == "false" ]] && unset par_csi
+[[ "$par_multiple" == "false" ]] && unset par_multiple
+
+samtools index \
+    "$par_input" \
+    ${par_csi:+-c} \
+    ${par_bai:+-b} \
+    ${par_min_shift:+-m "par_min_shift"} \
+    ${par_multiple:+-M} \
+    -o "$par_output"

src/samtools/samtools_index/test.sh

@@ -0,0 +1,91 @@
+#!/bin/bash
+
+test_dir="${meta_resources_dir}/test_data"
+
+echo ">>> Testing $meta_functionality_name"
+
+echo ">>> Generating BAM index"
+"$meta_executable" \
+  --input "$test_dir/a.sorted.bam" \
+  --bai \
+  --output "$test_dir/a.sorted.bam.bai"
+
+echo ">>> Check whether output exists"
+[ ! -f "$test_dir/a.sorted.bam.bai" ] && echo "File 'a.sorted.bam.bai' does not exist!" && exit 1
+
+echo ">>> Check whether output is empty"
+[ ! -s "$test_dir/a.sorted.bam.bai" ] && echo "File 'a.sorted.bam.bai' is empty!" && exit 1
+
+echo ">>> Check whether output is correct"
+diff "$test_dir/a.sorted.bam.bai" "$test_dir/a_ref.sorted.bam.bai" || \
+  (echo "File 'a.sorted.bam.bai' does not match expected output." && exit 1)
+
+rm "$test_dir/a.sorted.bam.bai"
+
+#################################################################################################
+
+echo ">>> Generating CSI index"
+"$meta_executable" \
+  --input "$test_dir/a.sorted.bam" \
+  --csi \
+  --output "$test_dir/a.sorted.bam.csi"
+
+echo ">>> Check whether output exists"
+[ ! -f "$test_dir/a.sorted.bam.csi" ] && echo "File 'a.sorted.bam.csi' does not exist!" && exit 1
+
+echo ">>> Check whether output is empty"
+[ ! -s "$test_dir/a.sorted.bam.csi" ] && echo "File 'a.sorted.bam.csi' is empty!" && exit 1
+
+echo ">>> Check whether output is correct"
+diff "$test_dir/a.sorted.bam.csi" "$test_dir/a_ref.sorted.bam.csi" || \
+  (echo "File 'a.sorted.bam.csi' does not match expected output." && exit 1)
+
+rm "$test_dir/a.sorted.bam.csi"
+
+#################################################################################################
+
+echo ">>> Generating bam index with -M option"
+"$meta_executable" \
+  --input "$test_dir/a.sorted.bam" \
+  --bai \
+  --output "$test_dir/a_multiple.sorted.bam.bai" \
+  --multiple
+
+echo ">>> Check whether output exists"
+[ ! -f "$test_dir/a_multiple.sorted.bam.bai" ] && echo "File 'a_multiple.sorted.bam.bai' does not exist!" && exit 1
+
+echo ">>> Check whether output is empty"
+[ ! -s "$test_dir/a_multiple.sorted.bam.bai" ] && echo "File 'a_multiple.sorted.bam.bai' is empty!" && exit 1
+
+echo ">>> Check whether output is correct"
+diff "$test_dir/a_multiple.sorted.bam.bai" "$test_dir/a_multiple_ref.sorted.bam.bai" || \
+  (echo "File 'a_multiple.sorted.bam.bai' does not match expected output." && exit 1)
+
+
+#################################################################################################
+
+echo ">>> Generating BAM index with -m option"
+
+"$meta_executable" \
+  --input "$test_dir/a.sorted.bam" \
+  --min_shift 4 \
+  --bai \
+  --output "$test_dir/a_4.sorted.bam.bai"
+
+echo ">>> Check whether output exists"
+[ ! -f "$test_dir/a_4.sorted.bam.bai" ] && echo "File 'a_4.sorted.bam.bai' does not exist!" && exit 1
+
+echo ">>> Check whether output is empty"
+[ ! -s "$test_dir/a_4.sorted.bam.bai" ] && echo "File 'a_4.sorted.bam.bai' is empty!" && exit 1
+
+echo ">>> Check whether output is correct"
+diff "$test_dir/a_4.sorted.bam.bai" "$test_dir/a_4_ref.sorted.bam.bai" || \
+  (echo "File 'a_4.sorted.bam.bai' does not match expected output." && exit 1)
+
+rm "$test_dir/a_4.sorted.bam.bai"
+
+#################################################################################################
+
+
+echo "All tests succeeded!"
+exit 0

src/samtools/samtools_index/test_data/script.sh

@@ -0,0 +1,12 @@
+#!/bin/bash
+
+# dowload test data from snakemake wrapper
+if [ ! -d /tmp/idxstats_source ]; then
+  git clone --depth 1 --single-branch --branch master https://github.com/snakemake/snakemake-wrappers.git /tmp/idxstats_source
+fi
+
+cp -r /tmp/idxstats_source/bio/samtools/idxstats/test/mapped/* src/samtools/idxstats/test_data
+# samtools index a_ref.sorted.bam -o a_ref.sorted.bam.bai
+# samtools index a_ref.sorted.bam -c a_ref.sorted.bam.csi
+
+