Using "proteomes" will pull in TrEMBL identifiers which won't get parsed in add_uniprot_annotations.py

[neely]

I recommended a student use the human reference proteome for the db to db blaster (https://github.com/pwilmart/PAW_BLAST), but that meant when it came to add annotations via add_uniprot_annotations.py, there were na's for TrEMBL entries.

An example protein in the 9606 reference proteome that will have na for gene symbol is SEPTIN3 A0A2R8Y4H2.

Now why two SEPTIN3 entries are in the reference proteome is beyond me (one sprot one trembl), and I don't believe this would be an issue if canonical was used for BLASTing. But still, if any TrEMBL entries come through in canonical, they won't get mapped in add_uniprot_annotations.py.

An easier answer is likely to change the keywlist_download.py to get TrEMBL, but that might be a huge file.

Not the expert here, B-)

[pwilmart]

Hi Ben, UniProt only has the extra annotation records for Swiss-Prot sequences. The DAT files (the flat text file) would be too large with all proteins. The DAT file at the FTP site is all species (that is about 580K for Swiss-Prot). You can populate some of the extra annotation fields by hand, if you want to. There may be a gene symbol in the description line, there are the accession and identifiers, and a hyperlink formula can be made. This is kind of a quick and dirty frosting on the cake tool. There may be some missed spots in the back. Cheers, Phil

…

On Nov 23, 2022, at 7:38 AM, Ben Neely ***@***.***> wrote: I recommended a student use the human reference proteome for the db to db blaster (https://github.com/pwilmart/PAW_BLAST), but that meant when it came to add annotations via add_uniprot_annotations.py, there were na's for TrEMBL entries. An example protein in the 9606 reference proteome that will have na for gene symbol is SEPTIN3 A0A2R8Y4H2 <https://www.uniprot.org/uniprotkb?query=proteome%3AUP000005640%20A0A2R8Y4H2>. Now why two SEPTIN3 entries are in the reference proteome is beyond me (one sprot one trembl), and I don't believe this would be an issue if canonical was used for BLASTing. But still, if any TrEMBL entries come through in canonical, they won't get mapped in add_uniprot_annotations.py. An easier answer is likely to change the keywlist_download.py to get TrEMBL, but that might be a huge file. Not the expert here, B-) — Reply to this email directly, view it on GitHub <#1>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFZNQST56CE7V7O5Y24USNDWJY3ATANCNFSM6AAAAAASJDHC6Y>. You are receiving this because you are subscribed to this thread.