DRAGEN Germline Enrichment ICA App

DRAGEN Germline Enrichment is an accurate and efficient end-to-end (FASTQ to VCF) secondary analysis solution for whole exome and targeted panel NGS data; it can be used for both germline and somatic variant calling. This app takes input files in FASTQ, ORA, BAM, and CRAM format. Files may be decompressed, go through Map/Align/Sort, and go through variant calling.

Before launching the app in ICA the corresponding bundle containing the pipeline must be linked to your project. Follow the steps on the ICA help site to link the DRAGEN bundle corresponding to the version and region you want to use (e.g. 'DRAGEN 4.4').

In-run PON support for CNV and Targeted Caller

Version 4.4.4 of the DRAGEN Germline Enrichment App introduced support for automatic generation of an in-run PON for both CNV calling and the Targeted Caller. An in-run PON is required for Targeted Caller from WES data (see Targeted Caller | Exome calling), and is the preferred replacement for a pre-built PON for CNV calling (see CNV Preprocessing | Panel of Normals).

Use the following steps to run the DRAGEN Germline Enrichment App using an in-run PON:

1. In your project go to 'Pipelines' and search for 'DRAGEN_Germline_Enrichment'. Click on the DRAGEN Germline Enrichment pipeline corresponding to the version you want to use.

2. Click 'Start Analysis'

3. Under 'General' section enter the name for the analysis under the 'User reference' field.

4. Complete the 'Pricing' section.

5. Select FASTQs/BAMs/CRAMs for all samples in the sequencing run. At least 5 samples are required for CNV and at least 30 samples are required for Targeted Caller.

6. Click '+' in the 'Reference' field and select the Illumina DRAGEN References version corresponding to the DRAGEN version being used. The correct reference version can be found on the DRAGEN Software Support Site page. For example, for DRAGEN 4.4 select 'Illumina DRAGEN v11 References'.

7. For human samples, select one of the supported pangenome reference TAR files containing 'graph' in the name and click 'Add'. Note that Targeted Caller only supports hg38, hg19 and hs37d5.

8. Click '+' in the 'Target BED File' field and select 'Illumina Enrichment BEDs' and then select the reference build selected previously.

9. Select the bed file corresponding to the library prep kit used. If the Illumina CS/PGx Custom Enrichment Research Panel was used, select 'Illumina_Custom_Enrichment_Panel_v2_DRAGEN_Targeted_Calling.<hg38|hg19|hs37d5>.bed'

10. Do not select files for the 'CNV Panel of Normals' or 'CNV Combined Counts' fields. An in-run PON will be built instead.

11. In the 'Settings | Variant Calling Options' section select the 'true' radio button for the 'Enable CNV calling' field.

12. Select the 'true' radio button for the 'CNV Enable In-Run Panel of Normals' field.

13. If the Illumina CS/PGx Custom Enrichment Research Panel was used go to the 'Settings | Additional Options' section and select the 'true' radio button for the 'Enable Targeted Calling' field.

14. In the field 'Excluded Sample for In-Run Panel of Normals' list the name of any samples from the sequencing run that should not be included in the PON. Examples of samples that can be listed here include:

    1. Samples that have not met QC requirements.

    2. Related samples from a large pedigree consisting of more than ~6% of the total number of PON samples (e.g. more than quad for PON of size 50).

    3. Samples that are known to not be copy-neutral in certain target regions of interest and would otherwise make up a large portion of the total PON samples.

15. The 'Samples Per Node' field defaults to 5. This value can be increased to reduce the cost overhead for spinning up more nodes, but will increase the total runtime. Decreasing this value will increase the cost overhead, but decrease the total runtime.