Update GATK to 4.6.0.0 in sv_pipeline_docker (#719)

broadinstitute · Aug 27, 2024 · 6b27c39 · 6b27c39
1 parent 5aa3b8c
commit 6b27c39
Show file tree

Hide file tree

Showing 13 changed files with 22 additions and 28 deletions.
diff --git a/dockerfiles/sv-base/Dockerfile b/dockerfiles/sv-base/Dockerfile
@@ -1,7 +1,7 @@
 # This is the base dockerfile for the GATK SV pipeline that adds R, a few R packages, and GATK
 ARG SAMTOOLS_CLOUD_IMAGE=samtools-cloud:latest
 ARG VIRTUAL_ENV_IMAGE=sv-base-virtual-env:latest
-ARG GATK_COMMIT="a33bf19dd3188af0af1bd17bce015eb20ba73227"
+ARG GATK_COMMIT="64348bc9750ebf6cc473ecb8c1ced3fc66f05488"
 ARG GATK_JAR="/opt/gatk.jar"
 ARG R_INSTALL_PATH=/opt/R
 
@@ -14,8 +14,8 @@ FROM $SAMTOOLS_CLOUD_IMAGE as samtools_cloud
 FROM $VIRTUAL_ENV_IMAGE as virtual_env_image
 RUN rm_unneeded_r_library_files.sh
 
-ARG GATK_BUILD_DEP="git git-lfs openjdk-8-jdk"
-ARG GATK_RUN_DEP="openjdk-8-jre-headless libgomp1"
+ARG GATK_BUILD_DEP="git git-lfs openjdk-17-jdk"
+ARG GATK_RUN_DEP="openjdk-17-jre-headless libgomp1"
 ARG GATK_COMMIT
 ARG GATK_JAR
 ARG DEBIAN_FRONTEND=noninteractive

diff --git a/src/svtk/svtk/pesr/pe_test.py b/src/svtk/svtk/pesr/pe_test.py
@@ -89,7 +89,8 @@ def _get_coords(pos, strand):
         startA, endA = _get_coords(record.pos, strandA)
         startB, endB = _get_coords(record.stop, strandB)
 
-        region = '{0}:{1}-{2}'.format(record.chrom, startA, endA)
+        # Add 1 because evidence is stored/indexed with 0-based coordinates
+        region = '{0}:{1}-{2}'.format(record.chrom, startA + 1, endA + 1)
 
         try:
             pairs = self.discfile.fetch(region=region, parser=pysam.asTuple())

diff --git a/src/svtk/svtk/pesr/sr_test.py b/src/svtk/svtk/pesr/sr_test.py
@@ -82,7 +82,7 @@ def test_record(self, record, called, background):
         # Clean up columns
         results['name'] = record.id
         results['bg_frac'] = results.called / \
-            (results.background + results.called)
+                             (results.background + results.called)
         results['bg_frac'] = results.bg_frac.fillna(0)
         cols = 'name coord pos log_pval called background bg_frac'.split()
 
@@ -120,7 +120,8 @@ def load_counts(self, chrom, pos, strand):
         """Load pandas DataFrame from tabixfile"""
 
         if pos > 0:
-            region = '{0}:{1}-{1}'.format(chrom, pos)
+            # Add 1 because evidence is stored/indexed with 0-based coordinates
+            region = '{0}:{1}-{1}'.format(chrom, pos + 1)
             try:
                 lines = self.countfile.fetch(region)
             except ValueError:

diff --git a/wdl/BAFTestChromosome.wdl b/wdl/BAFTestChromosome.wdl
@@ -113,17 +113,16 @@ task BAFTest {
       set -o pipefail
 
       java -Xmx~{java_mem_mb}M -jar ${GATK_JAR} PrintSVEvidence \
-        --skip-header \
         --sequence-dictionary ~{ref_dict} \
         --evidence-file ~{baf_metrics} \
         -L "${chrom}:${start}-${end}" \
         -O local.BAF.txt.gz
     else
       touch local.BAF.txt
       bgzip local.BAF.txt
+      tabix -0 -s1 -b2 -e2 local.BAF.txt.gz
     fi
 
-    tabix -s1 -b2 -e2 local.BAF.txt.gz
     svtk baf-test ~{bed} local.BAF.txt.gz --batch batch.key > ~{prefix}.metrics
 
   >>>

diff --git a/wdl/BatchEvidenceMerging.wdl b/wdl/BatchEvidenceMerging.wdl
@@ -158,7 +158,7 @@ task MergeEvidence {
     fi
 
     awk '/txt\.gz$/' evidence.list | while read fil; do
-      tabix -f -s1 -b2 -e2 $fil
+      tabix -f -0 -s1 -b2 -e2 $fil
     done
 
     /gatk/gatk --java-options "-Xmx~{java_heap_size_mb}m" PrintSVEvidence -F evidence.list --sample-names samples.list --sequence-dictionary ~{reference_dict} -O "~{batch}.~{evidence}.txt.gz"

diff --git a/wdl/GenotypeCpxCnvsPerBatch.wdl b/wdl/GenotypeCpxCnvsPerBatch.wdl
@@ -250,9 +250,9 @@ task RdTestGenotype {
     else
       touch local.RD.txt
       bgzip local.RD.txt
+      tabix -p bed local.RD.txt.gz
     fi
 
-    tabix -p bed local.RD.txt.gz
     tabix -p bed ~{bin_exclude}
 
     Rscript /opt/RdTest/RdTest.R \

diff --git a/wdl/MatrixQC.wdl b/wdl/MatrixQC.wdl
@@ -158,10 +158,9 @@ task PESRBAF_QC {
     else
       touch ~{print_ev_output}
       bgzip ~{print_ev_output}
+      tabix -f -0 -s 1 -b 2 -e 2 ~{print_ev_output}
     fi
 
-    tabix -f -s 1 -b 2 -e 2 ~{print_ev_output}
-
     /opt/sv-pipeline/00_preprocessing/misc_scripts/nonRD_matrix_QC.sh \
       -d ~{distance} \
       ~{print_ev_output} \
@@ -238,10 +237,9 @@ task RD_QC {
     else
       touch local.RD.txt
       bgzip local.RD.txt
+      tabix -f -p bed ~{print_ev_output}
     fi
 
-    tabix -f -p bed ~{print_ev_output}
-
     /opt/sv-pipeline/00_preprocessing/misc_scripts/RD_matrix_QC.sh \
       -d ~{distance} \
       ~{print_ev_output} \

diff --git a/wdl/PETestChromosome.wdl b/wdl/PETestChromosome.wdl
@@ -217,17 +217,16 @@ task PETest {
 
     if [ -s region.merged.bed ]; then
       java -Xmx~{java_mem_mb}M -jar ${GATK_JAR} PrintSVEvidence \
-        --skip-header \
         --sequence-dictionary ~{ref_dict} \
         --evidence-file ~{discfile} \
         -L region.merged.bed \
         -O local.PE.txt.gz
     else
       touch local.PE.txt
       bgzip local.PE.txt
+      tabix -0 -s1 -b2 -e2 local.PE.txt.gz
     fi
 
-    tabix -s1 -b2 -e2 local.PE.txt.gz
     svtk pe-test -o ~{window} ~{common_arg} --medianfile ~{medianfile} --samples ~{include_list} ~{vcf} local.PE.txt.gz ~{prefix}.stats
   >>>
   runtime {

diff --git a/wdl/RDTestChromosome.wdl b/wdl/RDTestChromosome.wdl
@@ -176,10 +176,9 @@ task RDTest {
     else
       touch local.RD.txt
       bgzip local.RD.txt
+      tabix -p bed local.RD.txt.gz
     fi
 
-    tabix -p bed local.RD.txt.gz
-
     Rscript /opt/RdTest/RdTest.R \
       -b ~{bed} \
       -n ~{prefix} \

diff --git a/wdl/ResolveCpxSv.wdl b/wdl/ResolveCpxSv.wdl
@@ -345,7 +345,6 @@ task ResolvePrep {
 
         if [ -s regions.bed ]; then
           java -Xmx~{java_mem_mb}M -jar ${GATK_JAR} PrintSVEvidence \
-                --skip-header \
                 --sequence-dictionary ~{ref_dict} \
                 --evidence-file $GS_PATH_TO_DISC_FILE \
                 -L regions.bed \
@@ -385,7 +384,7 @@ task ResolvePrep {
         > discfile.PE.txt.gz
     fi
 
-    tabix -s 1 -b 2 -e 2 -f discfile.PE.txt.gz
+    tabix -0 -s 1 -b 2 -e 2 -f discfile.PE.txt.gz
   >>>
 
   output {

diff --git a/wdl/SRTestChromosome.wdl b/wdl/SRTestChromosome.wdl
@@ -218,17 +218,16 @@ task SRTest {
 
     if [ -s region.merged.bed ]; then
       java -Xmx~{java_mem_mb}M -jar ${GATK_JAR} PrintSVEvidence \
-        --skip-header \
         --sequence-dictionary ~{ref_dict} \
         --evidence-file ~{splitfile} \
         -L region.merged.bed \
         -O local.SR.txt.gz
     else
       touch local.SR.txt
       bgzip local.SR.txt
+      tabix -0 -s1 -b2 -e2 local.SR.txt.gz
     fi
 
-    tabix -s1 -b2 -e2 local.SR.txt.gz
     svtk sr-test -w 50 --log ~{common_arg} --medianfile ~{medianfile} --samples ~{include_list} ~{vcf} local.SR.txt.gz ~{prefix}.stats
   >>>
   runtime {

diff --git a/wdl/SetSampleIdLegacy.wdl b/wdl/SetSampleIdLegacy.wdl
@@ -122,13 +122,13 @@ task SetSampleId {
     output_name="~{sample_name}.~{file_type}.txt.gz"
 
     if [ ! -f "~{evidence_file}.tbi" ]; then
-        tabix -s1 -b2 -e2 ~{evidence_file}
+        tabix -0 -s1 -b2 -e2 ~{evidence_file}
     fi
 
     mkfifo $fifo_name
     /gatk/gatk --java-options "-Xmx2000m" PrintSVEvidence -F ~{evidence_file} --sequence-dictionary ~{reference_dict} -O $fifo_name &
     awk '{$~{sample_column}="~{sample_name}"}' < $fifo_name | bgzip -c > $output_name
-    tabix -s1 -b2 -e2 $output_name
+    tabix -0 -s1 -b2 -e2 $output_name
 
   >>>
   runtime {

diff --git a/wdl/TasksGenotypeBatch.wdl b/wdl/TasksGenotypeBatch.wdl
@@ -344,10 +344,9 @@ task RDTestGenotype {
     else
       touch local.RD.txt
       bgzip local.RD.txt
+      tabix -p bed local.RD.txt.gz
     fi
 
-    tabix -p bed local.RD.txt.gz
-
     Rscript /opt/RdTest/RdTest.R \
       -b ~{bed} \
       -c local.RD.txt.gz \
@@ -435,9 +434,9 @@ task CountPE {
     else
       touch local.PE.txt
       bgzip local.PE.txt
+      tabix -0 -s1 -b2 -e2 local.PE.txt.gz
     fi
 
-    tabix -s1 -b2 -e2 local.PE.txt.gz
     svtk count-pe -s ~{write_lines(samples)} --medianfile ~{medianfile} ~{vcf} local.PE.txt.gz ~{prefix}.pe_counts.txt
     gzip ~{prefix}.pe_counts.txt
 
@@ -511,9 +510,9 @@ task CountSR {
     else
       touch local.SR.txt
       bgzip local.SR.txt
+      tabix -0 -s1 -b2 -e2 local.SR.txt.gz
     fi
 
-    tabix -s1 -b2 -e2 local.SR.txt.gz
     svtk count-sr -s ~{write_lines(samples)} --medianfile ~{medianfile} ~{vcf} local.SR.txt.gz ~{prefix}.sr_counts.txt
     /opt/sv-pipeline/04_variant_resolution/scripts/sum_SR.sh ~{prefix}.sr_counts.txt ~{prefix}.sr_sum.txt.gz
     gzip ~{prefix}.sr_counts.txt