Protein localization and children

[gocentral]

In attempting to annotate participation of two proteins as being responsible for localization of a third, I find I am having trouble with the curent definitions (or it could be lack of proper terms) for the following: POG is found in both nucleus and cytoplasm. Ggn1 is found in the cytoplasm; Ggn3 in the nucleolus. When co-expressed, POG/Ggn1 is exclusively in the cytoplasm; POG/Ggn3 in the nucelolus. Thus, Ggn1 and Ggn3 particpate in the establishment of the localization for POG.

I want to use GO:0008104 but the def indicates ", excluding that due to targeting to membrane bound subcellular organelles."

I think we need to modify this def. Clearly this is a protein localization event; if the membrane didn't exist, POG would still go wherever Ggn1 or Ggn2 would be located.

Reported by: hdrabkin

Original Ticket: "geneontology/ontology-requests/1946":https://sourceforge.net/p/geneontology/ontology-requests/1946

[gocentral]

Logged In: YES user_id=473796

Harold, did you see SF:1002541 - protein localization, targeting and transport? It's on a similar issue.

Personally I think that it would be better to have protein localization and targeting as children of protein transport, and define targeting as being anything involving a signal sequence and localization as non-signal- sequence-mediated. Something like this, maybe?

protein transport [i]protein localization -- ---[i]protein localization, signal sequence mediated ; syn:protein targeting

localization: transporting to and maintaining a protein in a specific area of the cell localization, signal seq mediated: transporting to and maintaining a protein in a specific area of the cell, mediated by a signal peptide sequence within the protein

That removes the organelle dependency problem we have at the moment.

Would that work for you? You could annotate to the protein localization term.

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=546388

Hi Amelia I'm a bit unconfortable with the "transport" part. For example, in this case, in the absence of the two proteins, the POG is distributed. Most likely a diffusion driven process; the protein in most likely small enough to get through the nuclear pore. When either of the other two are expressed, I think of the process like this: each of these two proteins have signals (such as NLSs), that determine where they will stay or be (one most likely gets to the nucleus by another protein recognizing the NLS), etc. They also bind to POG, and POG kind of gets stuck with whatever they do. Thus, the diffusion of POG would drive population to the nucleus, since once there, it is now bound, and the "free" concentration would drop. However, alternatively, the binding could occur in the cytoplasm directly after synthesis, and then POG would get dragged alont to the nucleus.

Let me give it a bit more thought, because, I can see that one can look at the process as "carrier driven", which has aspects of transport.

Original comment by: hdrabkin

[gocentral]

Logged In: YES user_id=546388

Ok.. as we define transport and transporter, I'm ok with your suggestion. The Oxford uses "membrane" too much in its definitiion of transport, and transporter (a membrane protein catalysing the pasage of molecules from one face of a membrane to another), etc. If we did this, then I don't think we could put "protein localization" under transport at all!

Original comment by: hdrabkin

[gocentral]

Logged In: YES user_id=473796

Actually, perhaps the structure should be

protein localization ; GO:0008104 [i]protein localization, signal sequence mediated ; GO:new ---[p]protein transport, signal sequence mediated ; GO:0006605 [p]protein transport ; GO:0015031 ---[i]protein transport, signal sequence mediated ; GO:0006605 [i]protein sequestering (?)

protein localization - alter existing term def: The processes involved transporting and maintaining a protein in a specific location. [can't really say 'in the cell' since transport includes extracellular transport, and presumably things can be localized outside a cell]

protein localization, signal seq mediated: The processes involved transporting and maintaining a protein in a specific location, mediated by a signal peptide sequence within the protein. could have the protein-xxx retention terms as children

protein transport, signal seq mediated - this would be the existing term 'protein targeting ; GO:0006605' with a slightly altered def: The directed movement of a protein into, out of, within or between cells, mediated by a signal peptide sequence within the protein. syn: protein targeting

We could also have a term to represent protein sequestering if desired, since we have various specific protein sequestering terms but no generic parent.

How does that look?

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=473796

Actually, perhaps the structure should be

protein localization ; GO:0008104 [i]protein localization, signal sequence mediated ; GO:new ---[p]protein transport, signal sequence mediated ; GO:0006605 [p]protein transport ; GO:0015031 ---[i]protein transport, signal sequence mediated ; GO:0006605 [i]protein sequestering (?)

protein localization - alter existing term def: The processes involved transporting and maintaining a protein in a specific location. [can't really say 'in the cell' since transport includes extracellular transport, and presumably things can be localized outside a cell]

protein localization, signal seq mediated: The processes involved transporting and maintaining a protein in a specific location, mediated by a signal peptide sequence within the protein. could have the protein-xxx retention terms as children

protein transport, signal seq mediated - this would be the existing term 'protein targeting ; GO:0006605' with a slightly altered def: The directed movement of a protein into, out of, within or between cells, mediated by a signal peptide sequence within the protein. syn: protein targeting

We could also have a term to represent protein sequestering if desired, since we have various specific protein sequestering terms but no generic parent.

How does that look?

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=516865

Your final suggestions seem OK (I like the clafification of the protein targeting terms). I also have a note on a scap of paper to request a protein seqeustering term ...I will thow this away.

I'm still not surre about the definition of localization.

"The processes involved transporting and maintaining a protein in a specific location" implies that it is broader than protein transport, not narrower ?

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=546388

Hi Amelia The def alterations will help. I notice that the word "transport" is causing consternation again! When I think of transport, I think membrane, so, protein transport sounds like transferring a protein across a membrane; of course, when we talk of localization, we can mean that, but not always.

Original comment by: hdrabkin

[gocentral]

Logged In: YES user_id=473796

Val - protein localization would be broader than transport with the new def and structure. I think that people use localization to refer both to moving something somewhere and also to keeping it in that place (eg. a protein localizes to the nucleus).

If people are in agreement about the most recent changes I proposed, I'll implement them.

wrt targeting, perhaps a similar distinction could be made between vesicle transport and vesicle targeting for the vesicle targeting terms under discussion in SF:1029238. The vesicle targeting terms talk about the protein coats which target vesicles to a location - but is there vesicle transport to those locations in non-coated vesicles? If there isn't, there doesn't seem to be a point in having the terms.

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=516865

>protein localization would be broader than transport with # ah didn't spot that, yes

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=473796

I have implemented the structure

physiological process [i]localization ; GO:0051179 ---[i]protein localization ------[i]protein transport ---[i]RNA localization ------[i]RNA transport ---[i]transport ------[i] (all the transport terms)

I haven't changed the defs of the targeting terms - I am going to do a new SF item and propose my Grand Unified Transport Theory in it, which will include the protein targeting stuff. The protein localization problem should be solved, though.

Original comment by: girlwithglasses

[gocentral]

• milestone: --> TermGenie
• assigned_to: nobody --> girlwithglasses
• status: open --> closed-fixed

Original comment by: girlwithglasses