2
$\begingroup$

I would like to retrieve a list of all the human genes and their proteins’ functional classification.

The Gene type attribute on ensembl.org is not specific enough, because it assign the label coding protein to all the proteins for which I am trying to classify. Also the KEGG Pathway and Enzyme ID option doesn't seem to help.

I have translated all the gene names to their GO representation, but now I am not sure how to query for their cellular function, biological process, and cellular component.

Is there such a platform that can retrieve this info from PANTHER?

Sadly, the reference proteome of humans doesn't contain a functional classification either:

 zcat UP000005640_9606.fasta.gz | grep -i mYOSIN-2
>sp|Q9UKX2|MYH2_HUMAN Myosin-2 OS=Homo sapiens OX=9606 GN=MYH2 PE=1 SV=1

# >db|UniqueIdentifier|EntryName ProteinName OS=OrganismName OX=OrganismIdentifier [GN=GeneName ]PE=ProteinExistence SV=SequenceVersion

Is there any way to use bgee, QuickGO, GONet, or Uniport to cluster genes using the customize columns?

EDIT:

I think that getting the most specific (GO term name) and least specific (GO domain) gene's GO values can be retrieved from ensmebl:

enter image description here

Improve this question
$\endgroup$
4
  • $\begingroup$ If you really need ALL genes then you won't find such a database, because we don't know all of their functions. $\endgroup$
    – Devon Ryan
    Commented Feb 15, 2020 at 8:31
  • $\begingroup$ Have you looked at GO annotations? They are an hierarchical annotation for gene function. For enzymes the EC numbers are also an option. PFam is for structural domains, which generally align with function. $\endgroup$
    – Matteo Ferla
    Commented Feb 15, 2020 at 13:05
  • 1
    $\begingroup$ GO terms are used routinely to cluster genes for enrichment analysis. There are three categories, location, function and process, and a gene/protein will have one or more descendants (it's a hierarchical ontology). There are online tools that classify genes, such as GOrilla or Reactome. Programmatically you can use say Goatools in Python. There is a R package (don't recall name). And I think a Julia. $\endgroup$
    – Matteo Ferla
    Commented Feb 16, 2020 at 11:32
  • $\begingroup$ @MatteoFerla thanks. Let me know if you have a working example. I found GOATOOLS: nature.com/articles/s41598-018-28948-z $\endgroup$
    – 0x90
    Commented Feb 16, 2020 at 16:14

0

Reset to default

Your Answer

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