Phenotype Ontologies Reconciliation Effort

Phenotype ontologies have emerged in the last decade as a means to organise and query phenotypic content. Representing ontologies in a common language, OWL, facilitated the integration of those ontologies into various tool-chains, such as terminological services (OLS, OntoBee), search engines and more recently sophisticated diagnostic and analytic tools such as Exomiser. The phenotype ontology community has made much progress to date standardising their development practices, in particular by committing to OBO principles. OBO principles encourage, for example, standard practices for representing identifiers, well-documented and open development practices and some shared standard vocabularies such as the Relation Ontology that facilitate integration.

Despite the wealth of shared practices and a well-developed shared ecosystem of tools such as Protege, Robot, owltools, and the OWL API, phenotype ontologies (and their cousins, disease ontologies) are often not very deeply integrated and interoperable. Entity-quality (EQ) patterns co-evolved with large phenotype ontologies such as the Human Phenotype Ontology (HP), Mammalian Phenotype Ontology (MP) and Zebrafish Phenotype Ontology (ZP) as a means to integrate ontologies without having to manually maintain links between them. EQ patterns allow ontologies to logically define phenotypes in terms of an entity (E, also called the 'bearer'), often an anatomical entity or a biological process, and a (modified) phenotypic quality (Q), such as 'abnormal morphology': For example, 'Abnormal eye morphology' could be defined as an 'abnormal morphology' that inheres in the 'eye'. By using standard vocabularies such as GO or UBERON for representing the entities and PATO for representing the phenotypic qualities, ontologies could now be combined and classified together.

While EQ patterns are now widespread in the phenotype community, their development was, up until now, mostly independent and despite frequent manual efforts to align definitions, large proportions of existing ontologies now contain logically incompatible definitions, which precludes smooth integration for cross-species queries and inference. The Phenotype Ontologies Reconciliation Effort is a community effort that attempts to reconcile logical definition across a number of important phenotype ontologies. The effort was formed as part of the POTATO (Phenotype Ontologies Traversing All The Organisms) Workshop, co-located with ICBO 2018.

Our whitepaper (and workshop report) outlines our objectives and rationale in more detail.

Current members

Ontology	People	Involved Groups	Status
Ascomycete Phenotype Ontology (APO)	Stacia Engel	Alliance of Genome Resources	IN
C. elegans Phenotype Ontology (WBPheno)	Chris Grove	WormBase	IN
Cellular Microscopy Phenotype Ontology (CMPO)	Simon Jupp	Samples, Phenotypes and Ontologies, EMBL-EBI	IN
Dictyostelium discoideum phenotype ontology (DDPHENO)	Petra Fey	dictyBase	IN
Drosophila Phenotype Ontology (DPO)	David Osumi-Sutherland, Clare Pilgrim	FlyBase	IN
Fission Yeast Phenotype Ontology (FYPO)	Midori A. Harris, Valerie Wood	PomBase	IN
Human Phenotype Ontology (HP)	Peter Robinson, Sebastian Koehler, Leigh Carmody	Monarch Initiative	IN
Mammalian Phenotype Ontology (MP)	Sue Bello	Alliance of Genome Resources	IN
Phenoscape Knowledge Base	Jim Balhoff, W. Dahdul	Phenoscape	IN
Planarian Phenotype Ontology (PLANP)	Sofia Robb	Stowers Institute for Medical Research	IN
Plant Trait Ontology (TO)	Laurel Cooper*, Marie-Angélique Laporte, Pankaj Jaiswal	Planteome, Bioversity	IN
Xenopus Phenotype Ontology (XPO)	Erik Segerdell	XenBase	IN
Zebrafish Phenotype Ontology (ZP)	Yvonne Bradford	ZFIN	IN

Joining

If you are interested in joining the phenotype ontology community, or need help with setting up your own phenotype ontology development environment, please send an email to the phenotype ontology editors mailing list.