assayed_using vs assayed_transcript for tor2 / SPBC216.07c

[kimrutherford]

In this case the extensions in the last three annotations are assayed_transcript but the others are assayed_using, all configured to display as affecting. So the summary isn't as compact as it could be:

(https://dev.pombase.org/gene/SPBC216.07c)

Note that assayed_using is configured to group genes into the same extension but assayed_transcript isn't, which is why the assayed_transcript annotations aren't combined with each other. That's a separate issue though.

[kimrutherford]

In order to help with this sort of problem, I've made a change in the dev site only so that if you mouse-over a relation display name it will show you the underlying relation in a pop-up. I've been meaning to do this for a while to help with debugging.

Example: https://dev.pombase.org/gene/SPBC216.07c

[ValWood]

the pop up is useful.

so we need to change the assayed using here to assayed transcript, but that is governed by the ontology term. Do you know why these are different?

[ValWood]

I guess we could have copy edited and that would keep the old relation. or edited manually in the text box and used "assayed using". In which case this problem could be quite pervasive.

Is it possible to make a log file of extension relations which don't match the config.

then we can decide whether to manually fix (if its only a handful), of have an automated fix.

[kimrutherford]

so we need to change the assayed using here to assayed transcript, but that is governed by the ontology term. Do you know why these are different?

I'm not sure. Sometimes we have changed what is configured in Canto but not updated all the extensions. That's most likely when we add a new configuration that just covers a sub-branch of the ontology.

The is the current configuration in Canto for FYPO:0001117:
FYPO:0000824 is_a assayed_transcript TranscriptID|SO:0000673 assayed transcript

Before May 2021 this was configured as assayed_using

Is it possible to make a log file of extension relations which don't match the config.

We can do that. It will several hours to implement though because it means handling ontologies. :-)