parentage? - translation factor activity

[gocentral]

Hi,

I'm looking at children of "nucleic acid binding", which are like this right now:

- nucleic acid binding (GO:3676) -- DNA binding (GO:3677) -- RNA binding (GO:3723) -- translation factor activity, nucleic acid binding (GO:8135)

Is there any form of translation factor activity that interacts with DNA? Considering that translation occurs at the ribosome with the mRNA and with tRNAs, it seems likely that this could be restricted to RNA binding.

The current def of 8135:

Functions during translation by binding nucleic acids during polypeptide synthesis at the ribosome.

would still be consistent with moving the term down to be a direct of 'RNA binding' instead of a direct child of 'nucleic acid binding', though possibly we'd want to alter the term name.

-Karen

Reported by: krchristie

Original Ticket: "geneontology/ontology-requests/4813":https://sourceforge.net/p/geneontology/ontology-requests/4813

[gocentral]

Logged In: YES user_id=516865 Originator: NO

Hi Karen,

I agree that no known translation factors bind to DNA

However, I'm confused by this

I have only used the translation factor terms 'translation iniitation factor' 'translation elongation factor' etc.

it now turns out these are children of translation factor activity, nucleic acid binding.

The definitions of these translation initiation factor etc do not mention that these need to bind nuceic acid in the def?. I couln't be certain that all of these bind RNA (I didn't know this parent term existed, i don't think it used to?)

It seems like these 2 terms should be merged, the nucleic acid binding parent removed, and if an RNA binding annotation is required it should be added as a seperate annotation.

Val

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=473890 Originator: YES

I don't really know whether this term is an appropriate parent for all of its child terms. Harold might have better comments in that area.

I was just observing that when looking at "nucleic acid binding", that it seems rather odd that this term is place at that level instead of under "RNA binding"

Looking at the GOID for "translation factor activity, nucleic acid binding" (GO:8135), and checking in the GO numbers file, I see that it was originally called "translation factor" and we now have this structure:

- translation regulator activity (GO:45182) -- translation factor activity, nucleic acid binding (GO:8135) -- translation factor activity, non-nucleic acid binding (GO:45183)

Most of the child terms of 8135 also have #'s in the 'original set' range, so this seems to be, at least partially, the result of renaming the 8135 term.

-Karen

Original comment by: krchristie

[gocentral]

Logged In: YES user_id=436423 Originator: NO

I am really not an expert in translation, and I agree that Harold can probably shed more light (I'll email him to have a look). But I see no problem with moving GO:8315 down to is_a RNA binding. It's possible that some of the children of GO:8135 should be moved up to is_a GO:45182.

m

Original comment by: mah11

[gocentral]

Logged In: YES user_id=546388 Originator: NO

Yes, it should be moved to RNA binding!

Original comment by: hdrabkin

[gocentral]

Logged In: YES user_id=473890 Originator: YES

Hey Harold, do you have thoughts on Val's questions about whether or not it is appropriate for all the child terms of 8135 to be children of 'nucleic acid binding'?

-Karen

Original comment by: krchristie

[gocentral]

Logged In: YES user_id=546388 Originator: NO

This isn't as straight forward as we'd like. Many of the activities listed are actually activities of a complex.

For example translation initiation factor activity

Not all initiation factors bind directly to the mRNA or tRNA . Of those that do, for example, eIF2, is a three-subunit factor. One of the subunits binds GTP, another binds the met-tRNAi.

So, for example, if I were annotating the alpha, beta, or gamma subunit of eIF2, it would contribute_to (argh) the initation factor activity, but

The beta subunit catalyzes the exchange of GDP for GTP, which recycles the EIF-2 complex for another round of initiation. The phosphorylation of eIF2a in mammals and yeast is a significant regulatory pathway that prevents the interaction of eIF2–GDP with eIF2B to exchange GDP for GTP and shuts down protein synthesis. The role of the gamma subunit is not very clear. The point is that what is being assayed as "initation factor activity" is usually the result of a very complex assay system (two assays are typically used, one measures formation of GTP-eIF2-met-tRNAi (ternary complex formation; filter binding using radiolabeled met(35S)-tRNAmeti with and without GTP), and a second measure formation of the pre-initation complex (which assumes the ternary complex has formed)

Another initiation factor, eIF2B, actually does not bind to RNA but recycles GDP-eIF2 to GTP-eIF2

Original comment by: hdrabkin

[gocentral]

Logged In: YES user_id=516865 Originator: NO

'translation factor activity' is a bit of an odd term, which biologists expect to see, but probably would not be a function term if we were being purist (i. e we would have complex and process terms and individual function ters like RNA binding and GTPase when known).Here I feel OK to use contributes_to with this proviso, becasue if you took the contributes_to away, nobody would really have a problem with these having the function annotation 'translation initiation factor'.

However, for the RNA binding I would rather annotate the individual gene products which are 'translation initiaion factors" to RNA binding if they have been shown to bind RNA directly, or have an RNA binding motif rather than hardwiring this into the ontology when it may not be true for all gene products which have this 'function term', even if the complex does bind RNA in its entirety.This would be a bit like annotating every splicosome associated gene product to RNA binding.

This would cause probelms for users who want to retrieve gene product which bind directly to RNA.

...my 2c...i mean p

Original comment by: ValWood

[gocentral]

Has there ever been a decision about what to do for this item? m

Original comment by: mah11

[gocentral]

I've moved GO:0008135 down to is_a RNA binding, since everyone has agreed on that all along. I've left everything else as is, since it hasn't caused big problems for some years.

To really deal with the issue of which-subunits-bind-RNA would require a major overhaul along similar lines as the transcription overhaul. I have no idea who may have time to do that, or when, so I've made an entry on the 'proposed content work' list and linked it to this item (even though I'm closing it because I've done the bit initially requested).

http://wiki.geneontology.org/index.php/Ontology\_Development\#Proposed\_content\_work

m

Original comment by: mah11

[gocentral]

• assigned_to: nobody --> gomidori
• status: open --> closed-fixed

Original comment by: mah11