Protocol for generating IGV/igv.js tracks for CNV annotations

List of generated tracks

• RefSeq Curated: ncbiRefSeqCurated.sorted.gtf.gz
• Gene-based annotations: ncbiRefSeqCurated.cover.annotated.vcf.gz
• ClinGen Prenatal (nstd75):
  • clingen.nstd75.GRCh37.variant_call.class1-2.vcf.gz
  • clingen.nstd75.GRCh37.variant_call.class3.vcf.gz
  • nstd75.GRCh37.variant_call.class4-5.vcf.gz
• ClinVar Clin SV (nstd102):
  • clingen.nstd102.GRCh37.variant_call.class1-2.vcf.gz
  • clingen.nstd102.GRCh37.variant_call.class3.vcf.gz
  • clingen.nstd102.GRCh37.variant_call.class4-5.vcf.gz
• ClinGen Curated regions:
  • ClinGen_region_curation_list_GRCh37.bed
• gnomAD:
  • all: gnomad_v2.1_sv.sites.bed.gz
  • controls only: gnomad_v2.1_sv.controls_only.sites.bed.gz
  • non-neuro: gnomad_v2.1_sv.nonneuro.sites.bed.gz
• DGV via UCSC: dgvGold.bb
• uniprot domains via UCSC: unipDomain.bb

---

Most of the tracks are downloaded from the respective sources, except for: - OMIM data that is supplied from gene-panel builder (to fetch directly once it is checked in) - OUSAMG data that we pass internally and process on TSD: - ousamg.class1_CG_aug2020_hg19.bed - ousamg.class4-5_hg19.pooled.bed

For igv.js all tracks needed to be - sorted - archived with bgzip - indexed with tabix

Current use of the pipeline:

# run the wrapper script to generate all tracks
make make-tracks

The created tracks will be in the sub-folder output.

Sub-commands for use in troubleshooting or further development:

# Build Docker
make build

# run the shell to get inside the container
make shell