Merge pull request #4 from TRON-Bioinformatics/fix-issue-multiallelic-filter

Fix issue with multiallelic filter

Author: priesgo

README.md

@@ -55,6 +55,8 @@ Optional input:
     * --skip_decompose_complex: flag indicating not to split complex variants (ie: MNVs and combinations of SNVs and indels)
     * --filter: specify the filter to apply if any (e.g.: PASS), only variants with this value will be kept
     * --input_bams: a tab-separated values file containing in each row the sample name, tumor and normal BAM files for annotation with Vafator
+    * --skip_multiallelic_filter: after VAFator annotations if any multiallelic variant is present (ie: two different 
+    mutations in the same position) only the highest VAF variant is kept unless this flag is passed
     * --snpeff_organism: the SnpEff organism name (eg: hg19, hg38, GRCh37.75, GRCh38.99)
     * --snpeff_datadir: the SnpEff data folder where the reference genomes were previously downloaded. Required if --snpeff_organism is provided
     * --snpeff_args: additional SnpEff arguments

modules/04_vafator.nf

@@ -45,7 +45,7 @@ process MULTIALLELIC_FILTER {
     tuple val(name), file(vcf)
 
     output:
-    tuple file("${vcf.baseName}.filtered_multiallelics.vcf"), emit: filtered_vcf
+    tuple val(name), file("${vcf.baseName}.filtered_multiallelics.vcf"), emit: filtered_vcf
 
     script:
     """

nextflow.config

@@ -27,7 +27,7 @@ env {
 // Capture exit codes from upstream processes when piping
 process.shell = ['/bin/bash', '-euo', 'pipefail']
 
-VERSION = '2.1.0'
+VERSION = '2.1.1'
 
 cleanup=true
 
@@ -63,6 +63,8 @@ Optional input:
     * --filter: specify a comma-separated list of filters to apply (e.g.: PASS,.), only variants with these values will be kept. If not provided all varianst are kept
     * --vcf-without-ad: indicate when the VCFs to normalize do not have the FORMAT/AD annotation
     * --input_bams: a tab-separated values file containing in each row the sample name, tumor and normal BAM files for annotation with Vafator
+    * --skip_multiallelic_filter: after VAFator annotations if any multiallelic variant is present (ie: two different
+    mutations in the same position) only the highest VAF variant is kept unless this flag is passed
     * --snpeff_organism: the SnpEff organism name (eg: hg19, hg38, GRCh37.75, GRCh38.99)
     * --snpeff_datadir: the SnpEff data folder where the reference genomes were previously downloaded. Required if --snpeff_organism is provided
     * --snpeff_args: additional SnpEff arguments

tests/run_test_8.sh

@@ -1,11 +1,22 @@
 #!/bin/bash
 
-### this test cannot be automated as it relies on SnpEff references which need to be downloaded beforehand
+
 source tests/assert.sh
 output_folder=output/test8
-snpeff_datadir=/home/you/snpeff
-nextflow main.nf -profile test,conda --output $output_folder --snpeff_organism hg19 --snpeff_datadir $snpeff_datadir
-test -s $output_folder/single_sample/single_sample.normalized.vcf || { echo "Missing test 1 output file!"; exit 1; }
-test -s $output_folder/tumor_normal/tumor_normal.normalized.vcf || { echo "Missing test 1 output file!"; exit 1; }
+echo -e "tumor_normal\tprimary:"`pwd`"/test_data/TESTX_S1_L001.bam" > test_data/test_bams.txt
+echo -e "tumor_normal\tnormal:"`pwd`"/test_data/TESTX_S1_L002.bam" >> test_data/test_bams.txt
+echo -e "single_sample\ttumor:"`pwd`"/test_data/TESTX_S1_L001.bam" >> test_data/test_bams.txt
+echo -e "single_sample\ttumor:"`pwd`"/test_data/TESTX_S1_L002.bam" >> test_data/test_bams.txt
+echo -e "single_sample\tnormal:"`pwd`"/test_data/TESTX_S1_L001.bam" >> test_data/test_bams.txt
+echo -e "single_sample\tnormal:"`pwd`"/test_data/TESTX_S1_L002.bam" >> test_data/test_bams.txt
+nextflow main.nf -profile test,conda --output $output_folder --input_bams test_data/test_bams.txt
+test -s $output_folder/single_sample/single_sample.normalized.vcf || { echo "Missing test 8 output file!"; exit 1; }
+test -s $output_folder/tumor_normal/tumor_normal.normalized.vcf || { echo "Missing test 8 output file!"; exit 1; }
+test -s $output_folder/single_sample/single_sample.normalized.vaf.vcf || { echo "Missing test 8 output file!"; exit 1; }
+test -s $output_folder/tumor_normal/tumor_normal.normalized.vaf.vcf || { echo "Missing test 8 output file!"; exit 1; }
 assert_eq `wc -l $output_folder/single_sample/single_sample.normalized.vcf | cut -d' ' -f 1` 53 "Wrong number of variants"
-assert_eq `wc -l $output_folder/tumor_normal/tumor_normal.normalized.vcf | cut -d' ' -f 1` 53 "Wrong number of variants"
+assert_eq `wc -l $output_folder/tumor_normal/tumor_normal.normalized.vcf | cut -d' ' -f 1` 53 "Wrong number of variants"
+assert_eq `wc -l $output_folder/single_sample/single_sample.normalized.vaf.vcf | cut -d' ' -f 1` 72 "Wrong number of variants"
+assert_eq `grep tumor_af $output_folder/single_sample/single_sample.normalized.vaf.vcf | wc -l | cut -d' ' -f 1` 35 "Wrong number of variants"
+assert_eq `grep normal_af $output_folder/single_sample/single_sample.normalized.vaf.vcf | wc -l | cut -d' ' -f 1` 35 "Wrong number of variants"
+assert_eq `wc -l $output_folder/tumor_normal/tumor_normal.normalized.vaf.vcf | cut -d' ' -f 1` 60 "Wrong number of variants"

tests/run_test_9.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+
+### this test cannot be automated as it relies on SnpEff references which need to be downloaded beforehand
+source tests/assert.sh
+output_folder=output/test8
+snpeff_datadir=/home/you/snpeff
+nextflow main.nf -profile test,conda --output $output_folder --snpeff_organism hg19 --snpeff_datadir $snpeff_datadir
+test -s $output_folder/single_sample/single_sample.normalized.vcf || { echo "Missing test 1 output file!"; exit 1; }
+test -s $output_folder/tumor_normal/tumor_normal.normalized.vcf || { echo "Missing test 1 output file!"; exit 1; }
+assert_eq `wc -l $output_folder/single_sample/single_sample.normalized.vcf | cut -d' ' -f 1` 53 "Wrong number of variants"
+assert_eq `wc -l $output_folder/tumor_normal/tumor_normal.normalized.vcf | cut -d' ' -f 1` 53 "Wrong number of variants"