FAQs

General Questions

• Does this pipeline support ORA compressed FASTQs?

  • Yes, all iterations of the pipeline support analysis from .fastq.ora files.
• Can I load more than one sample at a time?

  • Yes. When selecting biosamples, multiple can be selected as long as sex is set to auto-detect.

• Is there demo data available?

  • Yes, there is demo data in BaseSpace, ICA, and ICM.
  • To access the demo data on BaseSpace, go to the demo data tab and search for "5-Base".

    
  • On ICA, the demo data is included in the DRAGEN 4.4 bundle. After adding the DRAGEN 4.4 bundle to a project, navigate to the data tab, and find the "Illumina DRAGEN 5-Base Methylation Germline Demo Data" folder.
  • On ICM, create a study, and click "+ Add Data" and "Select from ICA project". Select "Bulk", "Methylation", and then "Illumina 5-Base Solution" as the format. Click "+ Add Demo Data", and navigate to the "Multiomics-Demo-Data" folder. Click into "Methylation", then "Illumina-5-base-solution", and import all samples in that folder.

  

  

  
• How does DRAGEN tell when a site is a genetic variant, or a methylated site?

  • A genetic variant impacts two strands of the DNA; a methylated site impacts only one.

    
• What DRAGEN features are 5-base aware?

  • Only the features and options specified in this documentation are formally supported. All other features should be considered experimental and used at risk.
  • For additional information, review the DRAGEN documentation: https://help.dragen.illumina.com/dragen-v4.5/product-guides/dragen-v4.5/dragen-methylation-pipeline/dragen-5base-pipeline
• How does DRAGEN 5-Base in Germline, Somatic, and Enrichment differ from DRAGEN Methylation?

  • In general, it's not recommended to use DRAGEN Methylation for 5-Base is because DRAGEN Methylation is meant for bisulfite data (unmethylated C -> T), and not our 5-Base solution (where you convert methylated C -> T). DRAGEN Methylation doesn't allow for variant calling with the 5-Base library. Additionally, the alignment isn’t 5-Base aware, and thus will be worse than when running DRAGEN Germline, Somatic, or Enrichment. However, there is a need to do a direct comparison between bisulfite data and 5-base, it is possible to run 5-Base libraries on DRAGEN Methylation.
• Should DRAGEN Germline, Somatic, and Enrichment be used for non-human samples?

  • Methylation calling for non-human samples is supported within DRAGEN Germline, Somatic, and Enrichment. Variant calling for non-human samples is not formally validated in these apps, but can be done experimentally. Some non-mammalian species have high CpH methylation; to support these samples, it's recommended to run --enable-cpg-methylated-mapping=false in the "Additional Arguments" section of the App. This setting is not recommended for analysis of samples where the majority of methylation occurs in a CpG context.
• How do you build a 5-base specific DRAGEN hashtable?

  • To build a 5-base specific DRAGEN hashtable, it's required to run --ht-methylated-cg=true . This will create a methyl_cg sub-directory, which the DRAGEN mapper will use automatically.
  • For additional information, please review the general DRAGEN documentation on non-human reference building: https://help.dragen.illumina.com/dragen-v4.5/product-guides/dragen-v4.5/dragen-reference-support/prepare-a-reference-genome#reference-seed-interval

Runtime and Cost Summary

• How long does the average analysis take and how much does it cost?

Pipeline	Num Samples	Coverage	Analysis Settings	Time	Compute Cost (iCredits)
DRAGEN Germline	1 sample	35x	Germline, 5-base methylation calling, SNV, CNV, SV	~1h	~5.5
DRAGEN Somatic	1 sample (tumor-only)	100x	Somatic, 5-base methylation calling, SNV	~3h15m	~12
DRAGEN Somatic	2 samples (tumor-normal pair)	100x/50x	Somatic, 5-base methylation calling, SNV	~4h45m	~35
DRAGEN Somatic	2 samples (tumor-normal pair)	100x/50x	Somatic, 5-base methylation calling, CNV, SV	~5h	~35
DRAGEN Somatic	2 samples (tumor-normal pair)	100x/50x	Somatic, 5-base methylation calling, SNV, CNV, SV	~6h30m	~40
DRAGEN Enrichment	1 sample	2500x	Somatic UMI Liquid, SNV	~30m	<1

Output Files Summary

• What output files does DRAGEN produce for 5-base data?

  The following table describes the key output files produced per sample:

  Output File	Filename Pattern	Description	Produced When
  BAM	{prefix}.bam (normal) / {prefix}_tumor.bam (tumor)	Aligned reads with methylation tags (XM, XR, XG). Contains methylation status for all MAPQ>0 proper-pair reads. Control reads marked as unmapped with ca:Z tag.	--enable-map-align-output true
  VCF	{prefix}.hard-filtered.vcf.gz	Small variant calls with integrated methylation reporting (M5mC fields). Reports methylation at ref and alt alleles. To learn more, review the VCF documentation.	--enable-variant-caller true
  gVCF	{prefix}.hard-filtered.gvcf.gz	Genome-wide variant calls including reference-confidence blocks with CpG methylation reporting. To learn more, review the gVCF documentation.	--enable-variant-caller true
  methyl_metrics	{prefix}.methyl_metrics.csv	Methylation calling QC: total Cs analyzed, methylation rates per context (CpG, CHG, CHH), strand alignment stats, plus lambda/pUC19 control metrics.	Always produced with --methylation-conversion illumina
  mapping_metrics	{prefix}.mapping_metrics.csv	Mapping QC: total reads, total bases, mapped %, Q30 %, duplicate %, MAPQ distribution, mismatch rates, soft-clip counts.	Always produced when mapping is enabled
  DRAGEN_report	{prefix}.dragen_report.json	Consolidated JSON metrics report aggregating all pipeline metrics into a single structured file for programmatic access.	Always produced (DRAGEN v4.5 default)
  CX_report	{prefix}.CX_report.txt.gz	Genome-wide per-cytosine methylation status (position, strand, meth/unmeth counts, context).	--methylation-generate-cytosine-report true (auto-enabled when no variant caller)
  M-bias	{prefix}.M-bias.txt	Per-read-position methylation bias for CpG/CHG/CHH in R1 and R2. Diagnoses end biases.	--methylation-generate-mbias-report true (default=true)

  Note: When --enable-variant-caller=true is set, the CX_report defaults to off (methylation is reported in the VCF/gVCF instead). Set --methylation-generate-cytosine-report true explicitly to produce both.

Output File Presence by Workflow

• Which output files are produced for each pipeline configuration?

  File	Germline WGS	Somatic TO WGS	Enrichment Germline	Enrichment Germline UMI	Enrichment Somatic TO
  BAM	Yes	Yes	Yes	Yes (collapsed)	Yes
  VCF (hard-filtered)	Yes	Yes	Yes	Yes	Yes
  gVCF	Yes	Yes	Yes	Yes	Yes
  CX_report.txt.gz	Yes	Yes	Yes	Yes	Yes
  methyl_metrics.csv	Yes	Yes	Yes	Yes	Yes
  mapping_metrics.csv	Yes	Yes	Yes	Yes	Yes
  CNV VCF	Yes	Yes	No	No	No
  SV VCF	Yes	Yes	Yes	Yes	Yes
  M-bias.txt	Yes	Yes	Yes	Yes	Yes
  umi_metrics.csv	No	No	No	Yes	No
  Nirvana annotations	Yes	Yes	Yes	Yes	Yes

Approximate Disk Usage Per Sample

• How much disk space does a single-sample DRAGEN 5-base run consume?

  The following table shows total disk usage per run (includes all output files plus the loaded reference hash table, which is ~80-90 GB):

  Configuration	Coverage	Output File Disk Used	Output Files Included
  Germline WGS	~35x	~136 GB	BAM, VCF, gVCF, CX_report, CNV VCF, SV VCF, annotated VCFs, methyl_metrics, mapping_metrics, m-bias, metrics JSON
  Somatic TO WGS	~100x	~296 GB	BAM, VCF, gVCF, CX_report, CNV VCF, SV VCF, annotated VCFs, methyl_metrics, mapping_metrics, m-bias, metrics JSON
  Enrichment Germline Panel (non-UMI)	~30x	~83 GB	BAM, VCF, gVCF, CX_report, SV VCF, annotated VCFs, methyl_metrics, mapping_metrics, m-bias, metrics JSON
  Enrichment Germline Panel UMI	~30x	~1 GB	BAM (collapsed), VCF, gVCF, CX_report, SV VCF, annotated VCFs, umi_metrics, methyl_metrics, mapping_metrics, m-bias, metrics JSON
  Enrichment Somatic TO Panel	~30x	~11 GB	BAM, VCF, gVCF, CX_report, SV VCF, annotated VCFs, methyl_metrics, mapping_metrics, m-bias, metrics JSON

  Note: Output file disk space does not includes the reference hash table (~80-90 GB).