Skip to content

workflows.md

Latest commit

 

History

History
134 lines (114 loc) · 3.55 KB

workflows.md

File metadata and controls

134 lines (114 loc) · 3.55 KB

Workflows

Tools & data

  • Reference
  • bwa
    • Nantes : version bwa-0.7.10
  • samtools
    • Nantes: version 1.1
  • bcftools
    • Nantes: version 1.1
  • picard
    • Nantes: version 1.126
  • gatk
    • Nantes: version 3.3.0
  • java
    • Nantes: version 1.7.0_25

Code

FASTQ Preprocessing

  • Nantes: none

Read Mapping Preprocessing

Nantes

Reads where aligned with bwa and option -M. Sample and Center have been specified in the -R header. BAM was sorted on the fly with samtools A typical mapping looks like:

 
bwa-0.7.10/bwa mem   -M \
	-R '@RG\tID:p11\tLB:sample_name\tSM:sample_name\tPL:ILLUMINA\tCN:center' \
	human_g1k_v37.fasta R1.fastq.gz R2.fastq.gz  |\
  samtools view -u -  |\
  samtools sort  -@ 3 -o sorted.bam -T tmp -

Merging the BAMs

Nantes

for a few samples (B00GG8Z, B00FX08,B00FX09,B00FX0A) , there were more than one pair of FASTQ. sorted bam where merged with picard:

java -XX:ParallelGCThreads=4 -jar /picard.jar MergeSamFiles \
   SO=coordinate \
   AS=true \
   MAX_RECORDS_IN_RAM=1000000 \
   CREATE_INDEX=false \
   COMPRESSION_LEVEL=9 \
   VALIDATION_STRINGENCY=SILENT \
   USE_THREADING=true \
   VERBOSITY=WARNING \
   O=merged.bam \
   TMP_DIR=tmp \
   COMMENT="Merged from (...) " \
    I=sorted1.bam   I=sorted2.bam

Removing/Flagging the duplicates

Nantes

PCR+ Optical duplicated duplicated were marked with picard MarkDuplicates:

java  -jar picard.jar MarkDuplicates \
	MAX_RECORDS_IN_RAM=1000000 \
	COMPRESSION_LEVEL=9 \
	TMP_DIR=tmp \
	INPUT/merged.bam \
	O=markdup.bam \
	M=markdup.metrics \
	AS=true PG=PMD \
	VALIDATION_STRINGENCY=SILENT \
	CREATE_INDEX=true

Realign Indels

Nantes

indels were realigned with GATK using the data (1000G_phase1.indels.b37.vcf and Mills_and_1000G_gold_standard ) provided by the GATK bundle in the regions of the capture+/-500pb.

java   -jar GenomeAnalysisTK.jar -T RealignerTargetCreator \
	-R human_g1k_v37.fasta \
	-L:capture,BED extended.capture.500.bed \
	-I markdup.bam \
	-o bam.intervals \
	--known:onekindels,VCF 1000G_phase1.indels.b37.vcf  \
	--known:millsindels,VCF Mills_and_1000G_gold_standard.indels.b37.vcf  \
	-S SILENT
java   -jar GenomeAnalysisTK.jar -T IndelRealigner \
	-R human_g1k_v37.fasta \
	-I markdup.bam \
  -L:capture,BED extended.capture.500.bed \
	-o realign.bam.tmp.bam \
	-targetIntervals  bam.intervals  \
	-known:onekindels,VCF 1000G_phase1.indels.b37.vcf  \
	-known:millsindels,VCF Mills_and_1000G_gold_standard.indels.b37.vcf  \
	-S SILENT

Base recalibration

Nantes

Bam were recalibrated with GATK using the data ( dbsnp138 ) provided by the GATK bundle in the regions of the capture+/-500pb.

java   -jar GenomeAnalysisTK.jar -T  BaseRecalibrator \
  -R human_g1k_v37.fasta \
	--validation_strictness LENIENT \
	-I realign.bam \
	-l INFO \
	-o  recal.grp \
	-knownSites:dbsnp138,VCF dbsnp_138.b37.vcf \
	-L:capture,BED extended.capture.500.bed \
	-cov ReadGroupCovariate \
	-cov QualityScoreCovariate \
	-cov CycleCovariate \
	-cov ContextCovariate
java   -jar GenomeAnalysisTK.jar -T PrintReads \
	-R human_g1k_v37.fasta \
	-BQSR recal.grp \
	-I realign.bam \
	-L:capture,BED extended.capture.500.bed \
	-o final.bam \
	--validation_strictness LENIENT \
	-l INFO