clade for GO terms?

[proost]

Analogous with gene families and InterPro domains, GO terms can also be assigned to a clade

Could easily be implemented.

[proost]

A reason not to do this is that lacking GO terms are often due to a lack of annotation. @mutwil ?

[mutwil]

I guess it's a somewhat interesting analysis, but I think only a handful of people would be interested in that. Skip if too complicated.