Pass job to germline docker call (resolves #478)

Author: jpfeil

src/toil_scripts/gatk_germline/common.py

@@ -22,11 +22,13 @@ def output_file_job(job, filename, file_id, output_dir, s3_key_path=None):
     work_dir = job.fileStore.getLocalTempDir()
     filepath = job.fileStore.readGlobalFile(file_id, os.path.join(work_dir, filename))
     if urlparse(output_dir).scheme == 's3':
-        s3am_upload(fpath=os.path.join(work_dir, filepath),
+        s3am_upload(job,
+                    fpath=os.path.join(work_dir, filepath),
                     s3_dir=output_dir,
                     s3_key_path=s3_key_path)
     elif os.path.exists(os.path.join(output_dir, filename)):
         job.fileStore.logToMaster("File already exists: {}".format(filename))
     else:
         mkdir_p(output_dir)
         copy_files([filepath], output_dir)
+

src/toil_scripts/gatk_germline/germline.py

@@ -566,6 +566,7 @@ def prepare_bam(job, uuid, url, config, paired_url=None, rg_line=None):
                                    config.g1k_indel,
                                    config.mills,
                                    config.dbsnp,
+                                   realign=False,    # Do not realign INDELs
                                    memory=config.xmx,
                                    cores=config.cores).encapsulate()
         sorted_bam.addChild(preprocess)
@@ -674,8 +675,9 @@ def setup_and_run_bwakit(job, uuid, url, rg_line, config, paired_url=None):
 
     return job.addFollowOnJobFn(run_bwakit,
                                 bwa_config,
-                                sort=False,         # BAM files are sorted later in the pipeline
+                                sort=False,             # BAM files are sorted later in the pipeline
                                 trim=config.trim,
+                                mark_secondary=True,    # Mark split alignments as secondary
                                 cores=config.cores,
                                 disk=bwakit_disk).rv()
 
@@ -739,7 +741,8 @@ def gatk_haplotype_caller(job,
 
     # Uses docker_call mock mode to replace output with hc_output file
     outputs = {'output.g.vcf': hc_output}
-    docker_call(work_dir=work_dir,
+    docker_call(job=job,
+                work_dir=work_dir,
                 env={'JAVA_OPTS': '-Djava.io.tmpdir=/data/ -Xmx{}'.format(job.memory)},
                 parameters=command,
                 tool='quay.io/ucsc_cgl/gatk:3.5--dba6dae49156168a909c43330350c6161dc7ecc2',

src/toil_scripts/gatk_germline/test/test_germline.py

@@ -214,7 +214,8 @@ def _get_default_inputs(self):
         inputs.ssec = None
         inputs.sorted = False
         inputs.cores = 4
-        inputs.xmx = '8G'
+        inputs.file_size = '0.5G'
+        inputs.xmx = '4G'
         inputs.output_dir = self.workdir
         inputs.suffix = ''
         inputs.unsafe_mode = False