Nexson specification of tree inference method

[josephwb]

The curator calls the tree inference method "Tree type". Nexsons call this "ot:curatedType".

The nexson property name should be changed to be more informative.

Posting this here since it involves multiple tools: the curator, UI, and validation.

[jar398]

Since it conveys the inference method, maybe the property should be called ot:inferenceMethod, and the curator app should present it as "Inference method". Just sayin'.

On Thu, Jan 15, 2015 at 5:13 PM, Joseph W. Brown notifications@github.com wrote:

The curator calls the tree inference method "Tree type". Nexsons https://github.com/OpenTreeOfLife/phylesystem-api/wiki/NexSON#table-i-predicate-vocabulary call this "ot:curatedType".

The nexson property name should be changed to be more informative.

Posting this here since it involves multiple tools: the curator, UI, and validation.

— Reply to this email directly or view it on GitHub https://github.com/OpenTreeOfLife/opentree/issues/552.

[mtholder]

Just chasing down the history: the earliest reference I found to this was in this thread: https://github.com/OpenTreeOfLife/phylografter/issues/86 the notes ( https://docs.google.com/document/d/1tVXWaq3kqctlCh3qonkXxCphLeEJi0_k-3kSx949sQE/edit#heading=h.wk263bxa5r6t ) from the Oct 9, 2013 which @pmidford alludes to don't specifically mention that tag, but do refer to a free-form text field describing the tree type.

edit: typo fixed (tread->thread)

[jimallman]

Since it conveys the inference method, maybe the property should be called ot:inferenceMethod, and the curator app should present it as "Inference method"

I think the problem is that it may contain other information, at least in the studies from phylografter. In OpenTreeOfLife/phylografter#86, @pmidford says it has historically held other information:

From Oct 9 call ot:curatedType (tree) - phylografter's type column - frequently contains inference method, but also other free text.

Maybe we should check for other values in the current phylesystem studies.

Hm, looks like the curation UI is actually inconsistent in its labeling. Here are a tree's properties when we're editing:

...and here's the same tree when we're just viewing the study (note the label for this field!):

I'll make a separate issue for this...

[pmidford]

Perhaps we should add ot:inferenceMethod, the curator UI seems pretty clear about what should be there. It would be nice to enforce a controlled vocabulary, but that is probably premature. Leave ot:type until someone can triage whatever else is in the free text for the phylografter studies.

On 1/19/15 9:40 AM, Jim Allman wrote:

Since it conveys the inference method, maybe the property should
be called ot:inferenceMethod, and the curator app should present
it as "Inference method"

I think the problem is that it /may/ contain other information, at least in the studies from phylografter. In OpenTreeOfLife/phylografter#86 https://github.com/OpenTreeOfLife/phylografter/issues/86, @pmidford https://github.com/pmidford says it has historically held other information:

From Oct 9 call
ot:curatedType (tree) - phylografter's type column - frequently
contains inference method, but also other free text.

Maybe we should check for other values in the current phylesystem studies.

Hm, looks like the curation UI is actually inconsistent in its labeling. Here are a tree's properties when we're editing:

screen shot 2015-01-19 at 9 34 46 am https://cloud.githubusercontent.com/assets/446375/5802002/c01339c2-9fbe-11e4-9408-69abb2975143.png

...and here's the same tree when we're just viewing the study (note the label for this field!):

screen shot 2015-01-19 at 9 35 04 am https://cloud.githubusercontent.com/assets/446375/5802004/c8a2eb3c-9fbe-11e4-88be-4875d647a35a.png

I'll make a separate issue for this...

— Reply to this email directly or view it on GitHub https://github.com/OpenTreeOfLife/opentree/issues/552#issuecomment-70502584.

[mtholder]

agreed that we need a controlled vocabulary. I ran

python scripts/phylesystem/ot_phylesystem_list_propery_values.py --report-ids  --tree '^ot:curatedType'

(on a recently updated version of a peyotl script). And retrieved the list of values for this field (and their associated trees) as shown in: https://docs.google.com/spreadsheets/d/1ki4mcSCNnl4z1uxUhCgENxCBFc104ua0hsn7UYU3jGw/edit#gid=100428982

it is a bit of a mess, as free text fields tend to be...

[josephwb]

From that list, it seems that ^ot:curatedType often contains information that should appear for the tree label. From the curator: In the nexson:

[kcranston]

I agree with Joseph about label. Having label be free text is fine (as a means to distinguish trees within studies) but that we want a controlled vocabulary on inference type. I would also like to start capturing information about tree type (gene tree, supermatrix, supertree, species tree), noting that 'tree type' might not be the best name for that field.

[jimallman]

I would also like to start capturing information about tree type (gene tree, supermatrix, supertree, species tree), noting that 'tree type' might not be the best name for that field.

Are these values mutually exclusive?

Is Tree type an optional choice, or should it be required?

[jimallman]

See related discussion in #555.