PMID search returns correct model but doesn't show the PMID in the table

[vanaukenk]

I tested searching for PMID:12186849.

The search returns four correct models, but when I open the last model in the list 'C. elegans PDI-2 pathway', I don't see that PMID listed as evidence for any of the annotations.

The other three models do show that PMID listed in the table.

[tmushayahama]

PMID:12186849 is there in the model,

below is a quick way I found out from the row data

Now as to why, these are the potentials. will debug soon

• The model has an error (2 part ofs), so one of them was read
• if the evidence was a causal edge linking activities
• the other triple is not really an activity
• A rare case is an orphan evidence. @kltm @vanaukenk some of the models have these dangling evidences, maybe we need to cleanup, if you go to graph editor and explode all out view

[tmushayahama]

Found the error

If you open the model in graph editor, this is where the PMID:12186849 is

the reason is the model has more than one part of on protein disulfide isomerase activity.

[vanaukenk]

Thanks @tmushayahama We should consult with @thomaspd wrt the two 'part of' relations. In theory, one MF could be a sub-function of another and curators could use the 'part of' relation between them, but we don't have that modeled in the Noctua form, yet.

[tmushayahama]

@vanaukenk @lpalbou This is still an issue. but this might be a ticket for landing page.

But the second part of the issue is with ART is that Noctua Form doesn't truly represent a noctua model. So in this case, do I replace what NF is seeing, has parsed or what the model has.

As an example let's say an reference on a "directly provides input for" or any edge or info NF doesn't read or sub functions

[vanaukenk]

@tmushayahama Let me review these models again and get back to you. We may need to split this out into separate tickets.