Recording confidence

[dosumis]

As I understand it, In dendrograms, confidence in any one split into subclusters is represented by line length - vertically in this representation:

e.g. In the above example, confidence in the sncg/VIP split is particularly low.

Is there some way we can get a score for this that we can add to the ontology? (short of hacking the layout figures in the dendrogram JSON). It might make sense to add this as an annotation edge between the two sibling classes).

[jeremymiller]

The confidence is stored in the "edgePar.conf" value in the json file. The "height" is also a meaningful value that has something to do with the gene expression distance. I'm not sure the precise details, but the shorter the height the more similar the cell types. Let me know if you need anything else.