phylotastic / phylotastic-portal

Web tool (ruby on rails) providing access to phylotastic services for acquiring species trees.
https://github.com/phylotastic/phylotastic/wiki
1 stars 3 forks source link

compose metadata stanza for scaling #314

Open arlin opened 6 years ago

arlin commented 6 years ago

The dev portal now shows scaled trees. However, the metadata report does not say anything about the scaling method. Instead it refers to "inference method" which says something like "induced_subtree from synthetic tree with ID opentree9.1". We need to extend this to include a description of the scaling, for the median, SDM and OT methods.

The resulting text could be something like "The induced_subtree from synthetic tree with ID opentree9.1 was scaled by the Supertree Distance Method (SDM) yada yada yada"

To do

LunaSare commented 6 years ago

Median inference method: Multiple chronograms with some or all species found in OToL tree store are summarized by calculating the median age of overlapping nodes into a distant matrix and then applying a clustering method such as UPGMA.

SDM inference method: Multiple chronograms with some or all species found in OToL tree store are summarized using the Supertree Distance Method (SDM)

OT method: The induced_subtree from synthetic tree with ID opentree9.1 scaled with OToL scaling service.

LunaSare commented 6 years ago

Where exactly goes the topology description?

arlin commented 6 years ago

@LunaSare, go to the dev portal, get a tree, then click on the "metadata" tab. This is the part that we need to fix. It has a slot for tree inference method but not for scaling.

This is because the metadata model is based on the MIAPA checklist (https://github.com/evoinfo/miapa/blob/master/checklist/MIAPA-checklist.md), which did not imagine this use-case. So, it does not have separate slots for "topology" and "scaling". It has slots for "tree inference method" and for "branch lengths". We can (1) create a slot for scaling; (2) put scaling methods in branch lengths; or (3) put scaling methods in "tree inference". Which is best?

Let's do the redesign first, then return to finish this issue. I have opened another ticket, #314, for a redesign. It would be great if you are available to brainstorm on that. The general idea is to document the trees.