exomiser
Add phenotype annotation sufficiency score to exomiser results.
And/or take this into account in the scores.
It seems many phenotypes match the same genes time and again due to the general phenotypes these are strongly associated with - e.g. NOS3, NOTCH1, BRAF, PEX5.
Scoring phenotypes not only against the entire profile but also seperate system branches would be really good, e.g. there may be cardio, pulmonary, craniofacial and limb phenotypes in a sample, yet because there are a lot of cardio terms associated with one gene this is probably artificially inflated.
Similarly colouring the matches by score would help to quickly determine how good a match really is when there are a lot of inexact matches.
We should add the sufficiency score meter to the Monarch website for sure.
Regarding the scoring we definitely need a replacement that takes into account the quality of the input phenotypes as well e.g. a case with a single ID HPO term can get loads of high-scoring matches but is not that informative. Do we have an issue defined for this already?
Do we have an issue defined for this already?
Isn't this the issue?! We need this feature still. It was reliant on this being ported into OwlSim3, but we could do something much simpler and check that there are at least three terms which have significant matches in the HP-HP mappings for example.