Open siiraa opened 6 years ago
paola Paola Roncaglia added a comment - 31-May-17 11:28
Hi Sirarat Sarntivijai,
You mentioned you might contact the SO people about this before we proceed. Not sure if they've already provided feedback or not. Thanks!
siiraa Sirarat Sarntivijai added a comment - 31-May-17 12:08
haven't done that. We discussed the EFO's views on whether we are seeing these as chemicals, or physical regions. Please consult with SO by filing on their trackers?
Paola Roncaglia added a comment - 31-May-17 13:54
Done: https://github.com/The-Sequence-Ontology/SO-Ontologies/issues/400
Hi Sirarat Sarntivijai and Simon Jupp
Karen Eilbeck from SO replied:
“We are working on a newer version of SO where the molecular entities are integrated with the top level of Chebi, and the corresponding sequence features are dependent continuants of some form. [at]mikebada is the person you should talk to. In my opinion, where you integrate these terms depends on what your intent is. Are you modelling the information carrying potential or the molecular/physical properties?”
I think that in EFO, when representing DNA vs RNA, the intent is to model based on molecular/physical properties, but when representing the various classes of RNA (ncRNA, snRNA etc.) the intent is to model based on information-carrying potential (from a chemical point of view, they’re all ribonucleic acids, though they may carry different modifications). It sounds like SO is trying to solve the same issues and when they do, I expect that EFO would benefit from mirroring the SO classification, DNA/RNA included and downwards. What do you think?
Meanwhile, I asked SO what their timeframe is.
Paola Roncaglia added a comment - 13-Jun-17 14:05
Update from Mike Bada:
"Our plan is for an ontology named the MSO (Molecular Sequence Ontology) composed of analogs of the large majority of current SO classes; this ontology is a representation of material entities (as well as their specific dependent continuants and some associated occurrents), and these are the sequence entity classes that are integrated within ChEBI. The SO then remains as an ontology of sequence entities represented as generic dependent continuants (GDCs) (as has been reported, though many/most of the actual definitions don’t actually seem to represent GDCs), and the SO GDC classes and can then be defined necessarily and sufficiently in terms of their corresponding MSO classes. I think we have a pretty good draft version of the MSO ready, but I’ll let Karen respond with her thoughts on a timetable. Lemme know if you have any more questions..."
Paola Roncaglia added a comment - 14-Jun-17 13:32
Waiting to hear from Mike Bada/Karen Eilbeck about timeframe to be able to use MSO.
paola Paola Roncaglia added a comment - 21-Jun-17 16:20
From Mike Bada at SO:
"We’ve put our draft MSO.owl on the SO GitHub site https://github.com/The-Sequence-Ontology/SO-Ontologies/ for you to take a look at. We recommend you use FaCT++ (which is included in Protege 5) to classify the ontology.
A quick overview: The MSO classes are integrated within high-level classes of the BFO and ChEBI ontologies. The analogs of the large majority of the SO classes (and the ones you’re most likely interested in) are under CHEBI:’chemical entity’, and more specifically, CHEBI:’molecular entity’ and CHEBI:group. Beneath the former, we’ve created a new fundamental class called sequence molecular entities, which encompass all ChEBI molecular entities that have sequences: nucleic acids, nucleic acid strands, and peptides. Beneath the latter are sequence molecular entity regions, which are segments of one or more residues as parts of sequence molecular entities. Sequence junctions and termini are for now placed under BFO:’immaterial entity’, and sequence attributes that we’ve kept are under BFO:’specifically dependent continuant’, and more specifically distributed between BFO:quality and BFO:disposition.
Hopefully this will help get you started. Let us know if you have any questions..."
Paola Roncaglia added a comment - 21-Jun-17 17:30
Karen Eilbeck at SO adds:
"Paola - I am working through it too. The IDs are a bit rough and I am compiling a list of questions/comments for Mike. One of the things to be aware is that the ontology contains both MSO and SO ids but they are all really MSO terms, in the final version, there would just be MSO IDS, and SO will be automatically generated."
Paola Roncaglia added a comment - 02-Aug-17 09:01
Note for self: just followed up on https://github.com/The-Sequence-Ontology/SO-Ontologies/issues/400 to ask if there's any update on stability of MSO and its readiness to be imported (or not).
Sirarat Sarntivijai added a comment - 03-Aug-17 14:20
Chris Mungall has also inquired about MSO maintenance plan on SO Tracker.
@paolaroncaglia Some pointer from David OS: Upper Ontology and Bridging axioms for GO https://github.com/geneontology/go-ontology/blob/master/src/ontology/extensions/go-bridge.owl
Upper = BFO Bridging axioms bridge GO, SO, CHEBI, PRO, NCBI_taxon:
Might be easier to understand what’s going on here by looking at go-plus.owl or the editor’s version of GO (best viewed by checking out go eds repo and using the catalog_xml in there to drive local imports.
We would like ti import SO terms, but these currently do not align with Chebi's definition of DNA and RNA molecule.
Need new branch in EFO for SO terms and axiomatise these to either DNA or RNA from Chebi