DAVID and Panther can't identify certain Uniprot Accessions

Question

[shpak.max]

I obtained a list of differentially expressed proteins from MS/MS data, and have been trying to use gene annotation tools like DAVID and Panther to get functional enrichment information, pathways, etc.

I've noticed that DAVID etc. are able to obtain data for Uniprot accessions that begin with P or Q, e.g. P12645, Q01538, etc, but when I include other valid Uniprot accessions, e.g. G3V3A0, K7EKX9, etc (basically anything other than P,Q followed by numeric) DAVID is unable to identify them.

Is there some reason these seemingly valid Uniprot accessions aren't identified by the annotation tools?

[GenoMax]

Likely because those (non P/Q) entries are computer-annotated (Trembl)/not manually reviewed.

[shpak.max]

What, if any, options do I have for running these accessions through the standard annotation tools?

[h.mon]

You can run these accession through UniProt Retrieve / ID mapping here.

[GenoMax]

Don't think so. That converter does not recognize trembl ID's.

[h.mon]

I tested with G3V3A0 and K7EKX9, and it worked.

[GenoMax]

What did you select as From and To options when it worked?

[h.mon]

The default

from: UniProtKB AC/ID  
to: UniProtKB

[GenoMax]

That conversion seems to return the same entry but does include a gene name. Would be curious to see if OP can use that otherwise ..

[h.mon]

After the conversion, you can click on columns and select several other info, for example, GO categories. Of course, it remains to be seen if these other columns will annotated.

Answer 1

[Elisabeth Gasteiger]

Both accession numbers, G3V3A0 and K7EKX9, are valid UniProtKB/TrEMBL accessions and are retrieved by the Batch Retrieve tool on the UniProt website (http://www.uniprot.org/uploadlists). DAVID seems to recognize them too, but I am not very familiar with the different options of the DAVID tool, and I recommend contacting their helpdesk.

Please note that accession numbers have no meaning, and you cannot deduce from the letters or the length of an accession number whether it identifies a Swiss-Prot (reviewed) or TrEMBL (unreviewed) entry (http://www.uniprot.org/help/accession_numbers). In particular, an entry keeps its accession number when it is reviewed and thus moves from TrEMBL into Swiss-Prot.

See also Does The First Letter Of A Uniprot Accession Number Have A Meaning? (several years old), and regarding the more recent addition of 10-character ACs to the previously used 6-character ACs:

The UniProtKB accession number format was extended in 2014 to allow for 10-character accession numbers: http://www.uniprot.org/help/2014/06/11/release

Like previously, the "new-style" accession numbers do not have any meaning either, except that they implicitly convey the information that the entry they identify has been created since June 2014. However, the reverse is not necessarily true, i.e. not all new entries have 10-character accessions. In any case, it is best not to interpret ACs in any way, but just consider them as unique identifiers for a UniProtKB record.