Is it safe to assume that a sequence assigned to a particular COG ID in an old version of eggNOG is related to sequences assigned to that same COG ID in the latest version?
In more detail, I would like to fund the ‘COG functional category’ for some eggNOG COG IDs retrieved from UniProtKB (e.g. www.uniprot.org/uniprot/F0YY04.txt) However, I don’t know the version of eggNOG that these COG IDs were created under. I am wondering if I can safely use the COG-functional mapping provided in the latest version of eggNOG (v4.5) to infer the function of my UniProt sequences.
I am aware that I could query the UniProt sequences against the latest version of eggNOG, but this would be much more time consuming than using the available mappings supplied by UniProtKB.