Open jmcmurry opened 9 years ago
cc: @DoctorBud @harryhoch thoughts?
ok. sounds interesting, @jmcmurry. It's almost as if you are suggesting that we rank each single phenotype in terms of similarity to a range of models? Certainly possible, but that might take some non-trivial precomputation...
I seem to recall we have had some discussions before about integrating the similarity score into browse portion of the app. However, I don't see any tickets open regarding this. I think it is somewhat unfortunate that someone who 1) browses phenotypes from the dovegraph and 2) finds one of interest ...is then unable to meaningfully rank results in the way that we already have working in phenogrid search mode.
I don't think phenogrid is necessarily the right tool for this limited use case (one row or one column); however, some similar look and feel (eg. consistent colors) might be a good idea.
I don't have time to flesh it out now, but wanted to write it down while I was thinking of it.
The browse and analyze pages do not have any links one to another and there's no consistency in how the results are ranked, displayed, grouped, or operated on.
http://beta.monarchinitiative.org/analyze/phenotypes/?input_items=HP%3A0000175&mode=search&target_species=all&limit=100
http://beta.monarchinitiative.org/phenotype/HP:0000175