-
To support interoperability, semantic artefacts should be designed based on well-supported patterns whenever relevant and possible. These patterns should be documented and published as a resource for …
-
Our web-based, prototype Phenotype Reconciliation Workbench [72] uses string matching and semantic criteria to identify ontology terms for potential alignment. We will refine this tool to improve usab…
-
```
Is the intent that OGMS follow principles of reuse, univocity?
homeostasis is in GO as a process. You're free to contest this, e.g. on the GO
tracker (I think it's a conventional view, and is us…
-
# `rBiopaxParser` is not reading in multiple Web Ontology Language (OWL) files
This is my first time working with `.owl` files; I need to work with both Experimental Factor Ontology (EFO) and The M…
-
What would be the best way to incorporate environmental data? Is it possible to link phenoscape and other semantic phenotype representations to ENVO?
-
- [x] Abstract
- [x] Simplify line “In our cells, proteins are constantly being created and are degrading, and are accumulating or interacting to produce the phenotypes that we see at a larger sca…
-
The plant trait ontology [TO](https://obofoundry.org/ontology/to.html) is the conceptual counterpart of OBA, but it contains (or will soon contain, as per @diatomsRcool) some phenotypey terms now, lik…
-
We will develop consensus guidelines for when a new disease should be defined. This will require review of mentions in the literature or other sources; determining whether molecular mechanisms would d…
-
Currently, the settings forms assumes the user will want to pull terms from a particular controlled vocabulary, or save a new term to that vocabulary. This is a huge assumption and because Tripal 3 en…
-
Required for #2. The current code is designed to run over the full set of published files and registry entries for all OBO projects, but we also need to be able to run for a single ontology's local fi…