Merge pull request #495 from nf-core/fix-indices-gatk

Add preserve realign + index

Author: jfy133

CHANGELOG.md

@@ -37,6 +37,7 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 * [#400](https://github.com/nf-core/eager/issues/400) -  Set correct recommended bwa mapping parameters from [Schubert et al. 2012](https://doi.org/10.1186/1471-2164-13-178)
 * [#410](https://github.com/nf-core/eager/issues/410) - Fixed nf-core/configs not being loaded properly
 * [#473](https://github.com/nf-core/eager/issues/473) - Fixed bug in sexdet_process on AWS
+* [#444](https://github.com/nf-core/eager/issues/444) - Provide option for preserving realigned bam + index
 * Increase MultiQC process memory requirements to ensure enough memory for large runs
 
 ### `Dependencies`

main.nf

@@ -2236,21 +2236,21 @@ if ( params.gatk_ug_jar != '' ) {
 
   output: 
   tuple samplename, libraryid, lane, seqtype, organism, strandedness, udg, file("*vcf.gz") into ch_ug_for_multivcfanalyzer,ch_ug_for_vcf2genome
-  tuple samplename, libraryid, lane, seqtype, organism, strandedness, udg, file("*realign.bam") optional true
+  tuple samplename, libraryid, lane, seqtype, organism, strandedness, udg, file("*realign.{bam,bai}") optional true
 
   script:
   defaultbasequalities = params.gatk_ug_defaultbasequalities == '' ? '' : " --defaultBaseQualities ${params.gatk_ug_defaultbasequalities}" 
-  keep_realign = params.gatk_ug_keep_realign_bam ? "T" : "F"
+  def keep_realign = params.gatk_ug_keep_realign_bam ? "" : "rm ${samplename}.realign.bam"
+  def index_realign = params.gatk_ug_keep_realign_bam ? "samtools index ${samplename}.realign.bam" : ""
   if (params.gatk_dbsnp == '')
     """
     samtools index -b ${bam}
     java -Xmx${task.memory.toGiga()}g -jar ${jar} -T RealignerTargetCreator -R ${fasta} -I ${bam} -nt ${task.cpus} -o ${samplename}.intervals ${defaultbasequalities}
     java -Xmx${task.memory.toGiga()}g -jar ${jar} -T IndelRealigner -R ${fasta} -I ${bam} -targetIntervals ${samplename}.intervals -o ${samplename}.realign.bam ${defaultbasequalities}
     java -Xmx${task.memory.toGiga()}g -jar ${jar} -T UnifiedGenotyper -R ${fasta} -I ${samplename}.realign.bam -o ${samplename}.unifiedgenotyper.vcf -nt ${task.cpus} --genotype_likelihoods_model ${params.gatk_ug_genotype_model} -stand_call_conf ${params.gatk_call_conf} --sample_ploidy ${params.gatk_ploidy} -dcov ${params.gatk_downsample} --output_mode ${params.gatk_ug_out_mode} ${defaultbasequalities}
     
-    if [[ ${keep_realign} == 'F' ]]; then
-      rm ${samplename}.realign.bam
-    fi
+    $keep_realign
+    $index_realign
     
     pigz -p ${task.cpus} ${samplename}.unifiedgenotyper.vcf
     """
@@ -2261,9 +2261,8 @@ if ( params.gatk_ug_jar != '' ) {
     java -jar ${jar} -T IndelRealigner -R ${fasta} -I ${bam} -targetIntervals ${samplenane}.intervals -o ${samplename}.realign.bam ${defaultbasequalities}
     java -jar ${jar} -T UnifiedGenotyper -R ${fasta} -I ${samplename}.realign.bam -o ${samplename}.unifiedgenotyper.vcf -nt ${task.cpus} --dbsnp ${params.gatk_dbsnp} --genotype_likelihoods_model ${params.gatk_ug_genotype_model} -stand_call_conf ${params.gatk_call_conf} --sample_ploidy ${params.gatk_ploidy} -dcov ${params.gatk_downsample} --output_mode ${params.gatk_ug_out_mode} ${defaultbasequalities}
     
-    if [[ ${keep_realign} == 'F' ]]; then
-      rm ${samplename}.realign.bam
-    fi
+    $keep_realign
+    $index_realign
     
     pigz -p ${task.cpus} ${samplename}.unifiedgenotyper.vcf
     """