GitHub release file details

We distribute pre-built gz files of transcripts via GitHub releases

The main files contain every historical transcript - so are useful for maximum compatibility when you want to resolve as many HGVSs as possible.

The gene symbol associated with a gene (and thus a transcript version) can change between GTFs (ie it is not part of what is frozen in a transcript version)

When generating these historical files, we use gene/transcript version from the most recent GTF

Annotation Consortium	Build	Example File
Ensembl	GRCh37	cdot-0.2.14.ensembl.grch37.json.gz
Ensembl	GRCh38	cdot-0.2.14.ensembl.grch38.json.gz
Ensembl	GRCh37/GRCh38	cdot-0.2.14.ensembl.grch37_grch38.json.gz
RefSeq	GRCh37	cdot-0.2.14.refseq.grch37.json.gz
RefSeq	GRCh38	cdot-0.2.14.refseq.grch38.json.gz
RefSeq	GRCh37 / GRCh38	cdot-0.2.14.refseq.grch37_grch38.json.gz

As files above contain the most recent transcript, if you want the transcript versions / gene symbols to match exactly what was in a release, you need to use .json.gz files produced from a single GTF release.

We host some of the versions used in VariantGrid, where they match the Release version used in a Ensembl VEP annotation, eg:

Annotation Consortium	Build	Example File	VEP release
RefSeq	GRCh37	cdot-0.2.14.GCF_000001405.25_GRCh37.p13_genomic.105.20201022.gff.json.gz	100-109
RefSeq	GRCh38	cdot-0.2.14.GCF_000001405.39_GRCh38.p13_genomic.109.20211119.gff.json.gz	-108