More Ensembl IDs than genes

Question

[patelk26]

Hello,

This is a very basic question, I am not sure if I understand Ensembl gene IDs. I have a gene quantification matrix and I can see there are around ~ 60K Ensembl gene IDs.

How can there be more (almost double) gene IDs than total number of genes in human genome? Can multiple gene IDs map to one gene? If yes, what is the purpose of having multiple gene IDs for one gene symbol?

[GenoMax]

If those ID's contain ENST* then they are alternate transcripts that are generated from a gene. There can be multiple such transcripts.

[patelk26]

I checked, they all begin from ENSG*

Answer 1

[Ram]

This is probably all genes (from total RNA), not just protein coding genes. You'd see 58-59K of these, as opposed to around 27K protein coding genes. The GENCODE GTF file is an useful resource to subset to just protein coding genes.

Can multiple gene IDs map to one gene? If yes, what is the purpose of having multiple gene IDs for one gene symbol?

I think this also happens. EnsEMBL gives freshly annotated genes new ENSG IDs. They'd share HGNC symbol with an existing ENSG entry, but might be present on a patch or an alt contig.

[patelk26]

Thank you for the explanation, this makes sense.