Ontologies¶

Biological ontologies such as the Sequence Ontology (SO) or the Gene Ontology are read from OBO files into an Ontology: a directed acyclic graph of OntologyTerm objects linked by is_a and part_of relationships.

from urllib.request import urlopen

from picea import Ontology

SO_URL = "https://raw.githubusercontent.com/The-Sequence-Ontology/SO-Ontologies/master/Ontology_Files/so.obo"
with urlopen(SO_URL) as response:
    so = Ontology.from_obo(string=response.read().decode())
len(so)

Terms are accessed by ID. OBO tags become attributes, with list values.

mrna = so["SO:0000234"]
mrna.name, mrna["def"]

(['mRNA'],
 ['"Messenger RNA is the intermediate molecule between DNA and protein. It includes UTR and coding sequences. It does not contain introns." [SO:ma]'])

parents and children are transitive: they contain all ancestors and all descendants of a term.

[(term.ID, term.name[0]) for term in mrna.parents]

[('SO:0000233', 'mature_transcript'),
 ('SO:0000673', 'transcript'),
 ('SO:0000831', 'gene_member_region'),
 ('SO:0001411', 'biological_region'),
 ('SO:0000001', 'region'),
 ('SO:0000110', 'sequence_feature')]

len(mrna.children)

Both return a new Ontology, so they can be filtered and grouped.

coding_rnas = mrna.children.filter(lambda term: "coding" in term.name[0])
sorted(term.name[0] for term in coding_rnas)

['coding_end',
 'coding_start',
 'distant_three_prime_recoding_signal',
 'five_prime_recoding_site',
 'flanking_three_prime_quadruplet_recoding_signal',
 'recoding_pseudoknot',
 'recoding_stimulatory_region',
 'three_prime_recoding_site',
 'three_prime_repeat_recoding_signal']