search

geneontology / go-annotation

This repository hosts the tracker for issues pertaining to GO annotations.
BSD 3-Clause "New" or "Revised" License
34 stars 10 forks source link

Protein complex representation in Noctua #1661

Open bmeldal opened 6 years ago

bmeldal commented 6 years ago

From GOC meeting in Cambridge, Oct 2017:

Spaceholder ticket. Please formulate the issue(s); make a proposal for next GOC mtg at NYU 2018.

bmeldal commented 6 years ago

Collecting use cases here: https://docs.google.com/document/d/1ZtAcjIyIQ_ycbuMHyvLA-KIJQtGenh82lxS-MKC6a_A/edit

bmeldal commented 6 years ago

@sylvainpoux can you add your demethylase case to the google doc, please? @ukemi I think you had examples, too?

sylvainpoux commented 6 years ago

Hi Birgit, I don't have the permission to edit the google doc

Please find the text here:

Case 3: N6-adenosine-methyltransferase

The N6-adenosine-methyltransferase complex is composed of 2 proteins in vertebrates: METTL3 (UniProt Q86U44) and METTL14 (UniProt Q9HCE5). Both METTL3 and METTL14 belong to the same methyltransferase family and at first glance METTL14 looks like a normal methyltransferase. However, the 3D-structures have shown that METTL3 constitutes the catalytic core (PubMed:27373337, PubMed:27281194, PubMed:27627798), while METTL14 constitutes the RNA-binding scaffold that recognizes the substrate.

The ability of METTL14 to have catalytic activity is unclear and a number of experimental evidences suggest that has no methyltransferase activity by itself (PubMed:27627798, PubMed:27281194, PubMed:27373337). According to some reports, has some methyltransferase activity in vitro (PubMed:24316715). However, other studies showed that METTL14 constitutes the RNA-binding scaffold that recognizes the substrate rather than the catalytic core (PubMed:27627798, PubMed:27281194, PubMed:27373337). 3D-structure studies showed that METTL14 contains a degenerate active site that is unable to accommodate donor and acceptor substrates (PubMed:27627798).

bmeldal commented 6 years ago

I've changed the permissions settings so anyone with the link can edit.

Have added your use case as well.

goodb commented 5 years ago

Hi @bmeldal . Work on converting pathways (especially from Reactome) into GO-CAMs has led to some discussion of how complexes are represented. It would be good if we can stay in alignment with your working group's decisions as they are made. Importantly we need to stabilize around a model for logically defining the relationships between complexes and the proteins and other entities that compose them. We've discussed one model here: https://github.com/geneontology/pathways2GO/issues/34#issuecomment-456606763 Would be great to have feedback on that idea.

bmeldal commented 5 years ago

Hi @goodb that's a very timely post :) It's discussed in https://github.com/geneontology/go-ontology/issues/14375 (relationships discussion) and https://github.com/geneontology/go-annotation/issues/1639 (RO discussion) and was on the agenda for today's postponed call so will be covered on the next call (date tbc). Please add your availability to the Doodle I sent out earlier and I'll try and find a date you can make.

Comments:

  1. We don't like contributes_to as it's a minefield: https://github.com/geneontology/go-annotation/issues/1650 (also on the next call agenda)

  2. Complexes can be used in the same way as any other GPs, esp if Complex Portal (CP) ACs are available (they are fully defined and species-specific).

  3. Reactome complexes are not always equivalent to CP complexes as the latter have specifically defined subunits and Reactome use sets. That's the main reason why we could never import Reactome complexes into CP as a batch.

  4. if logical definitions of complexes based on parts ever make it into the GO

Another minefield, see: https://github.com/geneontology/go-annotation/issues/1662

Plenty of scope for discussion! Speak soon.

srengel commented 5 years ago

do these discussions need to be kept in their own meeting (too many different meetings!) or can they be folded into the Annotation calls since these decisions affect all GO annotators?

bmeldal commented 5 years ago

This one could be done in a Noctua call. We decided to have separate calls for the complexes as we have a bunch of items to discuss and wanted to power through them before the GOC mtg as I can only attend the local ones and those in connection with other mtgs. Can you please add your availability to the Doodle and we'll see what we can do?

vanaukenk commented 5 years ago

I think the protein complexes working group still needs a separate meeting time to work through the items and issues they need to resolve. The outcome of those discussions will definitely be presented to the rest of the consortium for final feedback and approval. There are a lot of meetings, but for focused work like this, I think the separate calls are most productive. Anyone with a strong interest in protein complex representation in GO, is encouraged and welcome to attend.

goodb commented 5 years ago

Hi Birgit,

It seems there is a lot going on already :). I think it would be worthwhile to set aside time to specifically discuss Reactome + complexes + GO-CAMs (especially with people that have been involved in the conversion project like Peter, David, Kimberly and Huaiyu). But, I don't know where that fits on your overall priority list.

Just so we are on the same page. Our short term objective for the Reactome to GO-CAM project is to come up with a conversion that cleanly represents the Reactome knowledge in the GO-CAM structure in a way that, though it may potentially lose information, doesn't produce any false statements. We would then work from there to either improve the representation in Reactome (e.g. Peter is adding a lot of new GO annotations as a result of this work) or adapt the structure of the GO-CAMs such that they can capture any missing knowledge.

To do this effectively, we need decisions about how to represent Reactome complexes and Reactome Sets in GO-CAM models. Hence my poke to your thread :).

On Fri, Feb 8, 2019 at 8:32 AM vanaukenk notifications@github.com wrote:

I think the protein complexes working group still needs a separate meeting time to work through the items and issues they need to resolve. The outcome of those discussions will definitely be presented to the rest of the consortium for final feedback and approval. There are a lot of meetings, but for focused work like this, I think the separate calls are most productive. Anyone with a strong interest in protein complex representation in GO, is encouraged and welcome to attend.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/geneontology/go-annotation/issues/1661#issuecomment-461862205, or mute the thread https://github.com/notifications/unsubscribe-auth/AB_U6iqR5gqbKAAbDBci28g_W-4OFQVBks5vLaangaJpZM4P9sRm .

bmeldal commented 5 years ago

I had a look at the calendar and noticed you have weekly Reactome2GO calls. I can call in to one and we can see where I can help.

ukemi commented 5 years ago

This week's meeting will focus on the import of the BMP signaling pathway, but the issue of complexes is more general and we need to be sure that we are aligning with the decisions you are making in the working group. Perhaps I could schedule a future meeting to focus exclusively on complexes.

bmeldal commented 5 years ago

That would be good. I think the alignment of complexes and Reactome reactions is more specific to what we are doing in the Complex WG. We discussed this many times here at EBI and there are a whole lot of specific difficulties so it should be a bespoke meeting.