geneontology / noctua

Graph-based modeling environment for biology, including prototype editor and services
http://noctua.geneontology.org/
BSD 3-Clause "New" or "Revised" License
38 stars 12 forks source link

Ability to 'Add component' #419

Closed krchristie closed 7 years ago

krchristie commented 7 years ago

Hi,

Considering that there are options to "Add process" and "Add function", why is there no option to "Add component".

I need to add a component just as often as I need to add a process or a function.

It would be really great to be have a template for adding a component. It would be even more great, if the template had two [occurs in] boxes, since I frequently need to add both a cell type and an anatomical structure.

thanks,

-Karen

kltm commented 7 years ago

@thomaspd I feel like we had something like this (or at least planned to) at some point, but removed it due to either overhead considerations, or that it was slightly against the annoton-based mindset. Unfortunately, my notes yield little here; would you remember anything?

kltm commented 7 years ago

@krchristie Until there is a response from Paul, to explore this a bit more, let me try and get into the details of your use case a little. (pinging @ukemi or @vanaukenk or verification)

Within annoton modeling, cellular component is sometimes an "optional" addition. There may be cases where a CC term needs to be added separately, which would use the "Add individual" functionality. Do you have cases where CC terms are added without MF or BP; do these include an enabled by relation as well?

krchristie commented 7 years ago

It seems inconsistent that we templates for have 2 of the 3 aspects of GO, but not the third one. If you can just add an individual for anything that isn't an annoton, why do we have specific options for two of the aspects of GO, but not the third one?

I have lots of cases where I want to add CC info only. I come across many, many papers where all I get for some genes is CC info. Unfortunately (in my opinion) I am forced to link them to anonymous MFs. I can be OK with that, provided I can get my CC annotations out, however, frequently, I don't get out the basic annotations I want.

One of my current examples is this model: PMID-23525783-KRC. I am not getting any CC annotations to the term '9+2 motile cilium'. The annotations I want would include two pieces of extension information, one for the cell type and another for the anatomical term:

krchristie commented 7 years ago

Let me add, I would eventually also want to get similar CC annotations to '9+2 motile cilium' for the relevant genes from the model section that is centered around 'trachael epithelial cell' and 'trachea epithelium', but I've been experimenting with the model and don't think that side is connected properly (as of 5/8/17).

cmungall commented 7 years ago

Sorry if this is a bit cryptic, on a plane right now with slow connection don't have links to hand.

@dosumis has extended the yaml DOSDP framework we use for ontology design patterns to individuals. We should add some DPs for components following your use case:

vars:
  - component: "cell component"
  - cell: "cell"
  - structure: "anatomical structure"

triples:
  text: |
   :1 a %s .
   :2 a %s .
   :3 a %s .
   :1 part_of :2 .
   :2 part_of :3 .
|
  vars:
  - component
  - cell
  - structure

(this may not be the precise syntax we agreed on, don't have the ticket at hand)

DPs like this will drive the wizards in the left panel

cc @DoctorBud @balhoff

dosumis commented 7 years ago

classes:
   ...

relations:
   ...

vars:
  gene_product:  {however we specific gene product}
  function: "molecular_function"
  component: "cell component"
  cell: "cell"
  structure: "anatomical structure"

instance_graph:
    nodes: 
       cc: component
       c: cell
       a:  structure
       gp: gene_product

    edges:
      - [mf, enabled_by, gp]
      - [gp, occurs in cc]
      - [cc, part_of, c]
      - [c, part_of, a]

json schema spec: graph; triples spec

kltm commented 7 years ago

This will likely be swept up then in the conversion to the various template and table editors.