search

OpenKnowledgeSpace / methodsOntology

Repository for development of the neuroscience methods ontology.
MIT License
8 stars 9 forks source link

Why not use cell ontology? #13

Open mellybelly opened 8 years ago

mellybelly commented 8 years ago

https://github.com/NeuroscienceKnowledgeSpace/methodsOntology/blob/master/ttl/hbp_cell_ontology.ttl

Could pull content from the Cell Ontology, which is used as the logical basis of so many other ontologies including the Gene Ontology. They are happy to take requests and do a good job with attribution.

Will help with interoperability of your data down the road and reduce duplication of effort.

http://purl.obolibrary.org/obo/cl.owl

@dosumis @cmungall @nicolevasilevsky can help if you need.

tgbugs commented 8 years ago

Our current plan is to merge this with NIF-Cell which is in the process of receiving a complete overhaul. The modelling that we are doing for these entities is still quite unstable and likely to change as we start using it for tagging. In the long run we would love to get some of these classes merged into the cell ontology like we did in the past with SAO and NIF-GrossAnatomy.

cmungall commented 8 years ago

Are there long term plans for maintenance of NIF-Cell?

It may be faster to just merge NIF-Cell and the ontology in this repo into CL at around the same time.

Let us know when you plan to do this, I do periodic alignments with both NIF-Cell and the cell type portion of neurolex (which started off similar, but have divided, with neurolex pulling in stale copies of FBbt...). I was involved with the SAO->GO as well as NIFGA->Uberon, I have a good idea of how integration should proceed.

tgbugs commented 8 years ago

There are plans for NIF-Cell be a central part of the lexicon for the neuron entries in the neuroscience knowledge space. Hopefully those also translate into long term maintenance. We are also planning to integrate terms from the work here: https://github.com/renaud/neuroNER.

mellybelly commented 8 years ago

AFAIK there is no maintenance plan for NIF-Cell, and even if so, there is no reason not to utilize and contribute to unified community standards. The Cell ontology is used by very many large consortia such as ENCODE, FANTOM, the epigenomics road map, and the Gene Ontology. Why give up this interoperability?

tgbugs commented 8 years ago

At the moment I am the maintenance plan for NIF-Cell (for better or for worse).

The problem that I have encountered in the neuroscience community is that there is no settled standard for how to name cell types and much of the data and literature that uses even a consistent nomenclature is often missing explicit annotation of certain properties (eg brain region since it is often implied elsewhere in a paper). We can actually see this in the construction of hbp_cell_ontology which leaves out rat and somatosensory cortex since all the data that has been tagged with it is from rat somatosensory cortex.

Since we are still working to develop a nomenclature within the neuroscience community that minimally covers hbp data, allen brain data, @stripathy 's expression data, and entities from existing ontologies it seems like it might be premature to burden you all upstream during the process of hammering this out.

including @renaud for reference here, and cross referencing to https://github.com/renaud/neuroNER/issues/48