AssociatedTaxa to DynamicProperties

[seltmann]

@jhpoelen for record UCSB-IZC 00000557 the first pass interaction information was only in the associatedTaxa, second pass I structured it into dynamicProperties. Looks like it is not re-indexing in GloBI.

sourceLifeStageName=pupae ; sourceLifeStageID= ; experimentalConditionName=in nature ; experimentalConditionID=http://purl.obolibrary.org/obo/ENVO_01001226 ; interactionTypeName=inside ; interactionTypeId=http://purl.obolibrary.org/obo/RO_0001025 ; targetTaxonName=Mus ; targetTaxonId=https://www.gbif.org/species/2311167

[jhpoelen]

@seltmann thanks for sharing the details. GloBI was expecting dynamic properties in format [key]:[value] , using : as separator. Recent commit globalbioticinteractions/globalbioticinteractions@5255d0c adds support for [key]=[value] . The improvement should be available in next elton/GloBI version.

For this particular interaction, what do you expect GloBI to index? What interaction do you expect to show up?

[seltmann]

@jhpoelen I see. I was following the format in the Arctos/VertNet IPT which uses = and ;

I would expect:
sourceTaxon inside targetTaxon
targetTaxon interactsWith sourceTaxon (since not defined and inside not bidirectional and Sarcophagidae not parasites).

I am not sure what I expect with all of the added value data. Minimally it would appear in the data from the api. Although there are issues:

Arctos, for example, makes a statement experimental=no, I am making the statement experimentalConditionName=in nature. Are these both http://purl.obolibrary.org/obo/ENVO_01001226.

I do not know the answer here, but good example of best to use IDs!

[jhpoelen]

Ok.

How about the life stages?

[seltmann]

Well, this is where things get interesting. I could not find an ID for pupa, and the specimen itself is an adult, but it was reared from a dead mouse.

So pupa is important, because it is different than the lifeStage of the specimen.

[jhpoelen]

Two statement perhaps?

... adult emerged from dead mouse

... pupa lived inside dead mouse

and:

dead -> physiologicalStateName

[jhpoelen]

please see https://github.com/globalbioticinteractions/globalbioticinteractions/blob/master/eol-globi-taxon-resolver/src/main/resources/org/eol/globi/taxon/life-stage-mapping.csv#L56 for existing life stage mappings.

For more mappings (e.g., bodyPart) see https://github.com/globalbioticinteractions/globalbioticinteractions/tree/master/eol-globi-taxon-resolver/src/main/resources/org/eol/globi/taxon .

[seltmann]

Ok, so here is my "stuck point." I used the relationships emerged from in the tri-tropic and EOL created a namespace for this (according to: https://docs.google.com/spreadsheets/d/1GCChIIChXvf98oMFg7OKUznflZiSP_XLdI_6mm0IP2w/edit?usp=sharing). Problem is that http://eol.org/schema/terms/emergedFrom no longer exists? There is no link in RO for this relationship either. So, what are my options as a data provider. I can add the interaction in my data, but I cannot express it (linked) as I would like.

EOL: emerged from		http://eol.org/schema/terms/emergedFrom	emerged from;from;reared from				Specimen was determined to have been inside associate and came out from inside. Conditions around the specimen collection methods are unknown.

[jhpoelen]

@seltmann suggest to:

1. ensure that you submit a term request to RO ontology via http://obofoundry.org/ontology/ro.html
2. work with @jhammock of EOL and let her now that the emergedFrom term should be re-added.

[jhpoelen]

Note that emergedFrom term is mapped by GloBI to a more general term interactsWith see https://github.com/globalbioticinteractions/globalbioticinteractions/blob/c392536f70de860ace93d1d4d6bdc25c3f49ebad/eol-globi-lib/src/main/java/org/eol/globi/domain/InteractType.java#L135 .

Until 1. and 2. settle, I could promote to term to be supported by GloBI if it's urgent.