search

matsengrp / cft

Clonal family tree
5 stars 3 forks source link

Option to color nodes/branches by cluster merge step #137

Closed metasoarous closed 5 years ago

metasoarous commented 7 years ago

It just occurred to me that it would be nice looking at higher cluster steps if you could see where the naive of the prior seed cluster(s) is(are) on the tree. Coupled with coloring by the cluster step at which the merge happened, you could get a sense for how the cluster merging marches forward and affects the phylogenetic structure. Might be interesting... Thoughts?

metasoarous commented 7 years ago

Jotting down a quick thought on this; As we discussed in meeting, it doesn't really make sense to directly throw in the naive seqs from other clusterings, since they're not observed (nor inferred by the clustering under consideration). So throwing them into the phylogenetic analysis is affecting the rest of the phylogeny which definitely feels wrong. Perhaps though we could use something like pplacer that wouldn't mutate the tree, but only assigns a placement of the alternate naive sequences in the context of the tree in question. Would something like that make more sense? @matsen ?

metasoarous commented 7 years ago

I think I have the ticket here; when you're looking at a cluster step > 0, have a coloring option that lets you color branches/nodes by the step in which they joined. This should look good on #99, and is a lot simpler than my other ideas on this.