SPARK_Scripts

Scripts used in SPARK project For variant calling/de novo filtering/annotation

Variant Calling

Variant Calling pipeline based on GATK Best Practices v3.6 (https://software.broadinstitute.org/gatk/best-practices/workflow?id=11145). Exome-pipeline Dependencies are stored in WES_Pipeline_References.b37.biocluster.sh Functions are stored in exome.lib.sh

1. BWA&MarkDuplicates .fastq -> .bam

ExmAln.1a.Align_Fastq_to_Bam_with_BWAmem.sh (BWA V0.7.15; picard V2.7.1)

2. HaplotypeCaller .bam -> .g.vcf

ExmAln.2.HaplotypeCaller_GVCFmode.sh (GATK V3.6)

3. JointCall with GenotypeGVCFs .g.vcf -> vcf

ExmVC.1hc.GenotypeGVCFs.sh  (GATK V3.6)
ExmVC.2.MergeVCF.sh

4. Variant Quailty Recalibration with VQSR (this step seems already been removed from current version of best practices)

ExmVC.3.RecalibrateVariantQuality.sh (GATK V3.6)

5. Annotation with ANNOVAR

ExmVC.4.AnnotateVCF.sh

6. Some QC

ExmVC.5.MakeKinTestFilesFromVCF.sh (Kinship test, vcftools v0.1.15, PLINK1.9)
ExmAln.8.DepthofCoverage.sh (D15 vs Mean)
ExmQC.VCFsummary_stats.py (some variants statistics for vcf)

De novo Filtering

Exome-Filters

1. Seperate cohort VCF into trio

adhoc.1.vcf2trio.py

2. Filter by Genotype and other heuristic filtering

Get_Denovo.py #this script take trio vcfs and produce de novo variants according to filters stored in .yml files
DENOVO_FILTER.yml # More stringent filters
DENOVO_FILTER.Tier2.yml # Less stringent filters

3. IGV inspection

https://github.com/ShenLab/igv-classifier.git

ShenLab/SPARK_Scripts