id registry prototype

[Jonathan A Rees]

The node id registry prototype is starting to work. A writeup is in registry/README.md on the 'registry' branch of the reference-taxonomy repository, i.e.

https://github.com/OpenTreeOfLife/reference-taxonomy/blob/registry/registry/README.md

Maybe we can talk about this at tomorrow's hangout.

Jonathan

[Karen Cranston]

Are there any results that you can share?

--
You received this message because you are subscribed to the Google Groups "Open Tree of Life - Software Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to opentreeoflife-so...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

~~~~~~~~~~~~~~~~~~~~~~~
karen.c...@gmail.com
@kcranstn
~~~~~~~~~~~~~~~~~~~~~~~

[Jonathan A Rees]

What I have is that about .8% of plant genera in OTT 2.8, as identified by OTT id, require new registry ids in OTT 2.9, due to membership changes among the samples used to define the registry ids. Surprisingly, 43 of 351 of the version 2.8 registrations among those .8% actually resolve to taxa with different OTT ids (which almost certainly means different names), even though the original OTT id (and name) still exist in OTT 2.9.

I think these changes mostly reflect the activities of the NCBI curators, since 2.8 and 2.9 use the same versions of GBIF and IRMNG.

See the 'Example' section of the README.

Another measurement is that 337 plant registrations (in all) from OTT 2.9, e.g. Halimeda, become ambiguous (i.e. compatible with several nodes) when applied to the draft 4 synthetic tree. This means that the samples chosen based on taxonomy were not 'tight' enough to narrow in on the correct node in the synthetic tree. Either an exclusion that belongs to a sibling in the taxonomy does not belong to any sibling in the synthetic tree, or the samples, which come from different children in the taxonomy, all come from the same child in the synthetic tree. New, more specific registrations should be created to separate the ambiguity and the one that synthesis has identified as having the same membership as the taxon should be recorded as the replacement for the ambiguous one. This isn't in the README but I'll add it.