What Constitutes Different Ensembl Db Releases And How Can We Map Them To Ncbi Genome Builds

Question

[Prateek]

What are the differences between ensembl db releases for the same genome build?

For example ensembl's current release is 63 for the current human genome build is 37/hg19 but some of its previous releases (62, 61..) also correspond to human genome build 37).

How can we resolve issues of compatibility when it comes to querying an older ensembl release database for say, variants detected using the latest reference genome.

Thanks

Answer 1

[Bert Overduin]

Ensembl data is released on an approximately 3-month cycle. For some of the more obscure species the data between different releases can stay completely the same, but for the more popular species (e.g. human, mouse, zebrafish) a new release will always mean that there are new data. That can be a completely new gene set or an patched / updated gene set, new variants, new cross-references etc. etc. For each release we have release notes, that exactly tell you what is new in that particular release.

For most species a completely new gene set (or genebuild as we call it ourselves) is only created when there is a new genome assembly, but for e.g. human and mouse we aim to do this more often. For human for example we are using GRCh37 (hg19) from release 55 onwards with completely new genebuilds in release 55 and release 62. Apart from this gene sets are regularly patched / updated, for human this is almost every release the case. More information on this can be found on the Assembly and Genebuild page in the browser.