Reference proteome example: if we should replace A0A0D5BX39 with A0A1B7YQB5 in PHI-base 4

[Hsinyugithub]

This protein is in PHI-base 4. I think it is ok to do this. I would like to keep this as an example for future reference and would also like to know your opinion.

A0A0D5BX39 is going to be deleted from UniProt because Colletotrichum higginsianum (Taxon ID: 80884) is not part of a reference proteome. From the PHI‑base curation point of view, this UniProt accession could be replaced by A0A1B7YQB5 (Colletotrichum higginsianum (strain IMI 349063), Taxon ID: 759273), which is included in a reference proteome. These two protein sequences are 100% identical in alignment.

However, in the publication (PMID: 25992547), the GenBank ID KP180423 (protein_id = "AJW67919.1") points specifically to A0A0D5BX39.

In my investigation, in PHI‑base 4 there are at least 22 proteins annotated from Colletotrichum higginsianum (strain IMI 349063) (Taxon ID: 759273). such as

H1VA10
H1VKZ0
H1VQ06
A0A1B7YJY4
H1UV33
A0A1B7Y7A9
H1V442
A0A1B7Y872
H1UYP0
H1VJG6
A0A1B7XWH6
H1VPS8
H1VAN0
A0A4T0VYD8
A0A1B7Y5W1
H1VS96
H1VRI6
H1VGC5
H1V216
H1W5W6
H1VSH5
H1V111

There is only one from Colletotrichum higginsianum (Taxon ID
[80884]), which is A0A0D5BX39.

[jseager7]

I agree this one should be replaced.

Where any protein is scheduled for removal, and where the sequence is 100% identical to an existing accession on a reference proteome, I can't see any problem at all in replacing the old accession with the accession from the reference proteome.

I just hope for more cases where we can replace with 100% similarity 🙂