Use TAO ontology for annotations and integrate to PubAnnotations

[vemonet]

Choosing an ontology for text annotations

Potential ontologies:

• https://pubannotation.org using the TAO ontology: http://pubannotation.org/ontology/tao.owl
• https://github.com/dice-group/gerbil/wiki/NIF
• http://www.openannotation.org/spec/core too complex?

We chose to use the PubAnnotation format (simple but complete for our use-case): https://pubannotation.org

• API for PubAnnotation: http://www.pubannotation.org/docs/api/
• Source code: https://github.com/pubannotation/pubannotation
• Get a list of abstracts to annotate : https://pubannotation.org/docs/sourcedb/PubMed

To do

• Implement TAO ontology model in the Collaboratory to publish statements provenance complying with PubAnnotations specs
• Retrieve potential text to annotate from PubAnnotations in the Collaboratory. This could be implemented as a feature in PubAnnotation to get the texts which need annotations the most. e.g. get text from PubMed: http://pubannotation.org/docs/sourcedb/PubMed
• Retrieve nanopub annotations complying with PubAnnotations specs from the Nanopub network, then integrate the query to retrieve the nanopub annotations from the PubAnnotations REST API (could use grlc API to make it easier)

[vemonet]

@jdkim