0
$\begingroup$

In order to obtain orthologs for human genes, I am working with InParanoid version 7.0. This version of the database uses Ensembl Protein Identifiers (ENSP) from ENSEMBL version v54, which is based on NCBI 36 assembly of the human genome.

For my project, I need to have the Ensembl Gene Identifiers (ENSG) of these proteins, but in the last version of ENSEMBL, which is based on GRCh38 assembly. I need to use this version of InParanoid for consistency with previous analyzes.

Thus, my question is, could I just use biomaRt to map ENSP (from NCBI36) to ENSG (actual version) or this approach is inherently wrong?

Is it possible to "convert" ENSP identifiers from NCBI36 to GRCh38 and then do the map to ENSG?. I am aware of liftOver but I am not sure it would work here.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
4
$\begingroup$

Those IDs are elderly! Ensembl 54 was 2009!

I would recommend using BioMart combined with the ID history converter. The ID history converter will convert old IDs to new, and BioMart will convert ENSPs to ENSGs. You can either use the ID history converter with the ENSPs first, then the current BioMart to get the ENSGs. Or you can use Ensembl 54 BioMart to convert the old ENSPs to old ENSGs, then update the ENSGs with the ID history converter. I would probably do both and compare results.

Improve this answer
$\endgroup$
5
  • $\begingroup$ Totally agree! They are extremely old but I have to deal with them unfortunately. Thanks a lot for pointing me to the ID history converter, I wasn't aware about this tool. However, I think I didn't ask the question properly. Simplifying, could it be possible that a protein is ENSP00000387261 in v54 and then is ENSP00000387262 in v99? (Note that the example is hypothetical). I am not referring to the ENSPXXX.Y version of a protein but about if the identifier before the point could change between versions. $\endgroup$
    – plat
    Feb 4, 2020 at 15:59
  • $\begingroup$ Yes, they could have new identifier numbers, which is why you need to combine the ID converter tool with BioMart. Most will stay the same, some will be dead and some will be different. $\endgroup$
    – Emily_Ensembl
    Feb 4, 2020 at 16:14
  • $\begingroup$ Perfect. I am trying with the tool. However, I am having troubles with some proteins. For instance, ENSP00000259791. Apparently it is retired since the release 75, according to its ENSEMBL entry. However, if you introduce it in the ID history converter and look at the table it is still matched to the same ID in the 99 release. If you download the results file, this is not happening anymore. Why this contradiction?. Am I confusing something or is it a bug? Thanks a lot for your help @Emily_Ensembl $\endgroup$
    – plat
    Feb 6, 2020 at 14:01
  • $\begingroup$ That's weird. The tool should also give you three other IDs for that one, so you can go ahead with them. I will look into it. $\endgroup$
    – Emily_Ensembl
    Feb 6, 2020 at 14:09
  • $\begingroup$ Great. I am trying to parse the table for my 20,000 IDs in R so please let me know if you find what is the issue so I could take it into account systematically. Thanks! $\endgroup$
    – plat
    Feb 6, 2020 at 14:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.