Ontologies

Ontosunburst allow you to use multiple integrated ontologies or to import yours.

Integrated Ontologies

You can use integrated ontologies with the ontology parameter or --ontology and --onto arguments.

Ex:

ontosunburst --onto ec

ontosunburst(ontology='ec')

The different ontologies are :

• Metacyc classes : metacyc
• Enzyme commission numbers : ec
• Kegg classes : kegg
• Gene Ontology classes : go
• ChEBI roles : chebi

Ontologies can be used through integrated local files or through SPARQL requests.

Integrated local files

If you don't precise a path to your own ontology file, default integrated files will be used :

• metacyc : ontosunburst/Inputs/MetaCyc26_0_classes.json
• ec : ontosunburst/Inputs/enzymes_ontology.json
• kegg : ontosunburst/Inputs/kegg_onto.json

Personal file

If you don't want to use the default integrated files for the 3 previous ontologies you can give yours as input with the class_ontology parameter or --class_ontology and -cl arguments.

Ex :

ontosunburst --onto ec -cl path/to/my/ec/ontology/file

ontosunburst(ontology='ec', class_ontology='path/to/my/ec/ontology/file')

SPARQL requests (owl files)

For Chebi and Gene Ontology you must set up SPARQL servers and download OWL files to make queries.

You must launch the server before making each execution for these ontologies.

Requests are send by and URL linked to server. You can precise your own URL with the endpoint_url parameter or --url argument.

Ex:

ontosunburst --onto chebi --url my/chebi/url

ontosunburst(ontology='go', endpoint_url='my/go/url')

Or use the default if you don't fill the parameter :

• go : http://localhost:3030/go/
• chebi : http://localhost:3030/chebi/

Personal Ontology

If you want to make sunburst with an ontology not integrated in the package you can do it by generating your own files with the right format and passing them as arguments.

To do so ypu need to :

• Define a class ontology json file (or dictionary) in class_ontology parameter or --class_ontology and -cl arguments.
• Define the root of the ontology in root parameter or --root and -r arguments.
• Not fill (=None) the ontology parameter or --ontology and --onto arguments.

Class ontology dictionary

You must define all the concepts of your ontology and associate all their direct parents.

Ex :

CT_ONTO = {'a': ['ab'], 'b': ['ab'], 'c': ['cde', 'cf'], 'd': ['cde'], 
           'e': ['cde', 'eg'], 'f': ['cf'], 'g': ['gh', 'eg'], 
           'h': ['gh'], 'ab': ['ROOT'], 'cde': ['cdecf', 'cdeeg'], 
           'cf': ['cdecf'], 'eg': ['cdeeg', 'ROOT'], 'gh': ['ROOT'],
           'cdecf': ['ROOT'], 'cdeeg': ['cdeeg+'], 'cdeeg+': ['ROOT']}

You have also to precise the root of your ontology, here : 'ROOT'.

You can pass it at argument as the dictionary or write it in a .json file and pass it as argument as the path of the file.

If you use command line you can only give a path to a .json file.

Ex :

ontosunburst -cl path/to/json/file -r ROOT

ontosunburst(class_ontology=CT_ONTO, root='ROOT')
# OR
ontosunburst(class_ontology='path/to/json/file', root='ROOT')

You will obviously need to indicate at least your interest set to generate the associated sunburst (ex: ['a', 'b', 'c'])