Open ValWood opened 7 years ago
Hi @ValWood It looks like there are a number of IDA annotations to GO:0075522 'IRES-dependent viral translational initiation' for human proteins, so it would be good for the annotation group to review how we want to annotate the role of host proteins during viral infection. I took a quick look at our documentation and don't see specific guidelines for host proteins, but may be missing it. I forgot - do you have a Protein2GO account where you can flag these as annotation disputes? If not, we could forward this thread to Sylvain for comment, as I think some of these annotations originally came from SIB.
I think I remember from a lecture to students by a local virologist / gene expression expert, an assertion that the same mammalian IRES proteins co-opted by some viruses are also used by the mammal for responses to stresses like heat shock.
I remember thinking that when the way we did this changed from the pathogen branch to use the same terms that it could be a problem. I don't think I really realised what was happening fully until today.....
I have spotted a number of instances where the virus co-opts a host process, and the host gene is annotated to the viral usage, which is a process which exists in the host but the viral use may not match the host process . This could really cause problems with GO for analysing human data if it is heavily annotated to viral process usage......
I can't think of any way to prevent this other than to put back in a separate branch of GO.
It's very possible that viruses use the host's transcription factors, however it is not the role of the human protein to transcribe viral proteins, so I would say that these annotations are outside the score of GO.
On Tue, Jan 17, 2017 at 8:56 PM, Val Wood notifications@github.com wrote:
I remember thinking that when the way we did this changed from the pathogen branch to use the same terms that it could be a problem. I don't think I really realised what was happening fully until today.....
I have spotted a number of instances where the virus co-opts a host process, and the host gene is annotated to the viral usage, which is a process which exists in the host but the viral use may not match the host process . This could really cause problems with GO for analysing human data if it is heavily annotated to viral process usage......
I can't think of any way to prevent this other than to put back in a separate branch of GO.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/geneontology/go-annotation/issues/1507#issuecomment-273280967, or mute the thread https://github.com/notifications/unsubscribe-auth/AEj7ULyvhOlfO-UfIU5t9WOkNAB2uuAUks5rTRzlgaJpZM4LllgP .
Also human PCBP2. It is some type of RNA binding protein,
but it becomes annotated to IRES-dependent viral translational initiation and viral RNA genome replication
The RNA genome replication , although odd does not harm but IRES-dependent viral translational initiation will give a "translation" annotation.
I agree with the above discussion about host proteins and transcriptional/translational processes involving viruses. A word of caution. Consider that metazoa have evolved to deal with viruses since pre-neandrethal times, some host proteins most certainly evolved for antiviral purposes. Cytokines such as the Interferons are an example. Whatever decision is made, I hope all host proteins are not excluded for this reason.
Is it correct to annotation Humam/Mouse SSB to IRES-dependent viral translational initiation?
This isn't it's role in human, and results in it becoming annotated to "cytoplasmic" translation, when it is not involved in this process? (plus lots of propagation).
This doesn't seem to be a good way to annotate viral processes?