BP only template - issue with Table View

[krchristie]

When I enter a BP annotation via the BP only template, the BP term does not show up in the Table View, see: http://68.181.125.145:8910/workbench/simple-annoton-editor/?model_id=gomodel:5ac3d47600001521

[vanaukenk]

I just tried this using the acts upstream of or within relation, and I see an annotation to the root MF node with the evidence for the BP annotation, but no BP term listed. See the top line here with the annotation to fat-5:

I would also expect to see a line with the acts upstream of or within relation in the Relation (to MF) field and then the GO BP term with the same evidence.

@krchristie - is this consistent with what you are seeing?

[krchristie]

@vanaukenk - yes, I am seeing the same behavior. Apologies that the model url I pasted in went to the wrong model. I didn't have time to do screenshots, so I thought the link would be sufficient... I had used the 'START NEW MODEL' button and it looks like the url does not update to that of the new model. I have edited the original comment to go to the correct model now.

[tmushayahama]

@krchristie @vanaukenk can you check this plus its error message

[vanaukenk]

Thanks, @tmushayahama .
I now see both the root molecular function annotation as well as the BP in the summary view. However, I think we want to be consistent and make sure that the relation listed with the BP uses the correct relation between the MF and the BP (not between the gene product and the BP) which will be the same relation that is seen in the graph editor, e.g. causally upstream of or within. Here's a screenshot of what I see right now; the top annoton is from the default form and the bottom is from the BP only form:

[krchristie]

It looks like @tmushayahama has already responded to the previous comment by @vanaukenk to show the relationship between the MF and the BP, because I now see this:

which is consistent with the graph:

So, I think the SAE Table View now looks great, and matches the Graph. However:

1. I think the issue remaining here is that it is confusing to see the GP->BP relationships in the pulldown, but then get the MF->BP relationships in the display. Is there any reason to show the GP->BP relationships at all? (especially since we are not labelling the endpoints of the relationship to make it clear what this relationship is connecting). I think it might be better to just show the MF->BP relationships in the pulldown, so that what you choose is consistent with what you see in the Table View and in the Graph.

2. Following on from that we are trying to train people to use the MF->BP relationships, it seems that we might want to modify the BP ONLY template slightly to be clear about what is being stated. Right now, it is still saying "MF part of Biological Process" just like the DEFAULT template does, even though in this template, we are NOT saying that the (anonymous) MF is part_of the BP. Perhaps, it would be simple to put a "MF" label in front of the pull down to select the relationship on the first line, and then the label of the blank to select the BP would just need to say "Biological Process".

[vanaukenk]

Remove 'MF part of' text from the Biological Process line so as not to confuse the statement being made.

[krchristie]

Removing the MF part_of from this template and expanding it to include CL and anatomy extensions for this template looks great.

However, I continue to think that it is confusing to show one set of relationships, e.g. "acts upstream of or within" (GP -> BP, if I understand correctly) in the selection menu:

and then immediately convert to the other type of relationship, e.g. "causally upstream of or within" (MF -> BP) in both the Table View:

and in the graph editor:

especially when the template has added an anonymous MF, such that the statement being made in this annotation does actually seem to be about the presumed MF.

If what this template is doing is making relationships between the anonymous MF and the BP term, it seems it would be better to just put those relationships in the selection menu.

[tmushayahama]

First on this issue, I will fix the second occurs_in and should be part of, correct? @krchristie so this is about where the dropdown of the connections is located? Noctua 2.0 will address this issue. But where exactly can I put it?

Is it possible to ask a question before the dropdown. like Relationship to Anonymous MF? [____v]

[vanaukenk]

@tmushayahama
Thanks for fixing the second 'occurs in' issue; that relation should be 'part of'.

Wrt the MF - BP relations, I think what @krchristie is saying is that in the BP only version of the form it makes more sense to her to show root MF (molecular_function) and then display the MF - BP relations in the dropdown rather than the gene product-to-BP relations.

That display would then more accurately depict what is actually happening behind the scenes wrt creating the GO-CAM model, since the BP only form automatically generates an annotation to root MF.

We can discuss tomorrow how best to address this going forward.

[vanaukenk]

@balhoff - we need to be clear on how to do the pass through from, say, an MF directly to an Anatomical Entity.

[vanaukenk]

For BP only in form2.0:

With respect to, the relations between root MF and BP on the BP only form, change to 'causally upstream of or within' and children.

Also, show default annotation to root MF.