Changeset 5840

Timestamp:

Feb 25, 2020, 9:35:02 AM (3 years ago)

Author:

Nicklas Nordborg

Message:

References #1218: Implement MIPs alignment

Added functions for re-header, split and mark duplicates.

File:

• other/pipeline/trunk/mips_functions.sh (modified) (1 diff)

other/pipeline/trunk/mips_functions.sh

@@ -245 +245 @@
 }
 
+# Re-header the BAM file
+# Replace SQ entries in the novo.bam header with reference genome dict
+# - The novoalign genome index was built from a genome with ambiguity codes
+#   and therefore the md5 checksums don't match the standard reference
+#   genome.
+# - Mismatching md5 values in the SQ entries will cause both picard and
+#   GATK tools to crash downstream
+# Global parameters:
+#   $samtools: Path to samtools executable
+#   $GenomeDict: Path to genome .dict file
+function reheader_bam {
+  ${samtools} view -H tmp/novo.bam | grep -v "^@SQ" > tmp/header.sam
+  grep "^@SQ" ${GenomeDict} >> tmp/header.sam
+  ${samtools} reheader -P tmp/header.sam tmp/novo.bam > tmp/novo_reheaded.bam
+}
+
+# Split novo_reheader.bam into seperate files for concordant, discordant and unmapped read pairs
+# Global parameters:
+#   $samtools: Path to samtools executable
+function split_bam {
+  ${samtools} view -b -h -@ 2 -f 3 tmp/novo_reheaded.bam > tmp/concordant.bam
+  ${samtools} view -b -h -@ 2 -G 12 -F 2 tmp/novo_reheaded.bam > tmp/discordant.bam
+  ${samtools} view -b -h -@ 2 -f 12 tmp/novo_reheaded.bam > out/unmapped.bam
+}
+
+# Mark duplicates with Picard that is aware of UMIs
+# Parameters:
+#   $1: Prefix to .bam file to process
+# Global parameters:
+#   $MarkDuplicatesOptions: Options to Picard
+function mark_duplicates {
+  local prefix=$1
+  
+  ./stdwrap.sh ./picard2 UmiAwareMarkDuplicatesWithMateCigar \
+    ${MarkDuplicatesOptions} \
+    -UMI_METRICS out/${prefix}.umi_metrics.txt \
+    -METRICS_FILE out/${prefix}.dedup_metrics.txt \
+    -INPUT tmp/${prefix}.bam \
+    -OUTPUT out/${prefix}.bam \
+    >> markduplicates.out
+}
+