DRAGEN – Variant Calling and Filtering

DRAGEN functional equivalence discovery and filtering of SNPs and indels.

This workflow uses the GATK HaplotypeCaller for SNP and indel discovery according to the GATK Best Practices. When the workflow runs in DRAGEN mode, it produces a Dragstr model that is used during variant calling, and it performs hard filtering.

This workflow is maintained by the Broad Institute and is written in Workflow Description Language (WDL). Further documentation can be found here.

Workflow Inputs

The workflow requires sample and reference information. The user may decide whether or not to run the pipeline in DRAGEN functional equivalence mode by setting the value of the run_dragen_mode_variant_calling input.

The Broad Institute provides various test inputs hosted in GCP that can be used to run the pipeline.

Input	Description
`calling_interval_list`	Interval list used for variant calling
`evaluation_interval_list`	File containing the target set of genomic intervals
`haplotype_scatter_count`	Scatter count used for variant calling
`break_bands_at_multiples_of`	Breaks reference bands at genomic positions that are multiples of this number; used to reduce (g)VCF file size
`input_bam`	Input BAM
`input_bam_index`	Input BAM index
`ref_fasta`	Reference fasta
`ref_fasta_index`	Reference fasta index
`ref_dict`	Reference dictionary
`dbsnp_vcf`	dbSNP VCF file
`dbsnp_vcf_index`	dbSNP VCF file index
`base_file_name`	String used for output files; can be set to a read group ID
`final_vcf_base_name`	Base name for the output (g)VCF file; can be set to a read group ID
`agg_preemptible_tries`	Number of preemptible machine tries
Optional inputs	Set `run_dragen_mode_variant_calling` to true and `use_spanning_event_genotyping` to false to run in DRAGEN functional equivalence mode.
`run_dragen_mode_variant_calling`	Boolean used to indicate that DRAGEN mode should be used for variant calling; must be true to compose DRAGstr model and perform variant calling with HaplotypeCaller in dragen-mode [false]
`use_spanning_event_genotyping`	Boolean used to call the HaplotypeCaller –disable-spanning-event-genotyping parameter; default set to true so that variant calling includes spanning events. Set to false to run the DRAGEN pipeline [true]
`contamination`	Optional estimated contamination for the variant calling tasks
`ref_str`	Optional STR location table used for DragSTR model auto-calibration; see ComposeSTRTableFile for more information
`make_gvcf`	Boolean that indicates if workflow should produce a GVCF or VCF file [true]
`make_bamout`	Boolean that indicates if workflow should sort and gather the BAM files into one file [false]
`use_gatk3_haplotype_caller`	Use the GATK3.5 HaplotypeCaller [false]
`skip_reblocking`	Skip reblocking (read more in the Reblocking section) [false]
`use dragen_hard_filtering`	Boolean that indicates if workflow should perform hard filtering using the GATK VariantFiltration tool with the –filter-name “DRAGENHardQUAL” [false]

Workflow Outputs

The key outputs of the workflow are variant calls in either VCF or gVCF format. Various metrics and associated files are also produced.

Output	Description
`CollectVariantCallingMetrics`	Outputs from calling variants with HaplotypeCaller.
`vcf_summary_metrics`	(g)VCF summary metrics
`vcf_detail_metrics`	(g)VCF detail metrics
`output_vcf`	Final (g)VCF
`output_vcf_index`	Index for the final (g)VCF
`bamout`	Optional corrected and merged BAM if make_bamout is true
`bamout_index`	Optional index for the corrected and merged BAM if make_bamout is true
`MergeBamOuts`	Output from sorting and merging the BAM files, then correcting the merged BAM file.
`output_bam`	Output aligned recalibrated BAM if provided_output_bam is true
`output_bam_index`	Optional index for the aligned recalibrated BAM if provided_output_bam is true

References

Reference data hosted in GCP may be found here.

Containers

Containers used by the pipeline are hosted in the Broad Institute’s public container registry, and the public biocontainers registry in quay.io.