search

matsengrp / olmsted

B-cell repertoire and clonal family tree explorer
http://olmstedviz.org
GNU Affero General Public License v3.0
15 stars 4 forks source link

Don't show cluster multiplicities for seed lineage pruned trees #57

Open metasoarous opened 6 years ago

metasoarous commented 6 years ago

As pointed out in #11, this isn't defined unless we actually have a minadcl tree. At the very least for these trees we would select the regular multiplicity by default, but preferably, cluster_multiplicity wouldn't even be an option in this case. Will have to do the same for multiplicities and cluster_multiplicities pie charts (see #56).

eharkins commented 6 years ago

Does that mean that currently, when we show the cluster_multiplicity for seeded families, we're actually showing the minadcl tree, otherwise there would be no cluster_multiplicity since that metadata is a product of minadcl? If so, this seems relevant to #6

metasoarous commented 6 years ago

I think it may be a bit random, depending on hash order or something. Remember that you're getting at all this data through the tripl relational querying system.

This is really part of the purview of #6, as part of that issue needs to be the UI around selection of which tree you want to look at.

metasoarous commented 6 years ago

Right now it looks like we're instead just showing the multiplicity for each, which is probably better than showing cluster_multiplicity for each, since this would be empty for the seed tree when selected (see #6). But ideally, we would by default switch to cluster_multiplicity when we show the minadcl tree.