This extended abstract presents an ongoing work by the Pistoia Alliance Ontologies Mapping project to develop user requirements for an ontologies mapping service.
The Pistoia Alliance Ontologies Mapping project1 was set up to nd or create better tools or services for mapping between ontologies in the same domain and to establish best practices for ontology management in the Life Sciences. It was proposed through the Pistoia Alliance Ideas Portfolio Platform2 which was selected by the Pistoia Alliance Operations Team for development of a formal business case.
The project has delivered a set of guidelines for best practice which build on existing standards. We show how these guidelines can be used as a \checklist" to support the application and mapping of source ontologies in the disease and phenotype domain. These guidelines are accessible on a public wiki.3. The project has contributed a process to evaluate existing ontologies mapping tools and a new phenotype track at OAEI (OM) 2016 to support evaluation of ontology matching (OM) algorithms.
User Requirements and Evaluation of existing
Ontologies Mapping Tools Another important output of this project was to specify the requirements for an Ontologies Mapping Tool. The three major aspects of the user requirements are illustrated in Figure 1. This shows the speci cation of the 1) User interface; 2) Framework, including ontology matching algorithms and 3) Import and export of mappings.
These requirements were used in a preliminary survey that established that such tools already exist which substantially meet them. Therefore, we have developed a formal process to de ne and submit a request for information (RFI) from existing ontologies mapping tool providers. We will summarise our ndings from this evaluation of seven ontologies mapping tools from academic and commercial providers on the poster. The guidelines and RFI materials, including the speci c details of the requirements are accessible on the same public wiki, mentioned previously. 3
Evaluation of Ontology Matching Algorithms for Disease and Phenotype A critical component of any Ontologies Mapping tool is the embedded ontology matching algorithm. Therefore, the Pistoia Alliance Ontologies Mapping 1 http://www.pistoiaalliance.org/projects/ontologies-mapping 2 IP3:https://www.qmarkets.org/live/pistoia/home 3 https://pistoiaalliance.atlassian.net/wiki/display/PUB/Ontologies+
Mapping+Resources Project is supporting development and evaluation of ontology matching algorithms though sponsorship and organisation of the new Disease and Phenotype track for OAEI (OM) 2016,4 which is also be summarised in this section. This new track has been organised because currently, mappings between ontologies in a given data domain are mostly curated by bioinformatics and disease experts in academia or industry, who would bene t from automation of their procedures. This could be accomplished through implementation of ontology matching algorithms into their existing work ow environment or investment in an ontologies mapping tool for management of the ontologies mapping life cycle. 3.1
There will be two tasks for this novel OAEI track which will comprise of pairwise alignment of:5 { Human Phenotype Ontology6 (HPO) to Mammalian Phenotype Ontology7 (MP), and { Human Disease Ontology8 (DOID) to the Orphanet and Rare Diseases Ontology9 (ORDO).
The rst task is important for translational science where human phenotype includes inherited diseases and mammalian phenotype ontology which is based on rodents as a model mammalian organism for many laboratory studies, including gene knock out. The second task includes representation of rare human diseases in both ontologies which are of fundamental importance for understanding how genetic variation can cause disease.
We have extracted a \baseline" reference alignments for the track based
on the available BioPortal mappings [
The evaluation of the Disease and Phenotype Track will be run with support of
the SEALS infrastructure.11 Systems will be evaluated and ranked according to
the following criteria:
4 The OAEI is an annual campaign for the evaluation of ontology matching systems [
This poster describes access to ontologies guidelines and a process to evaluate existing tools and algorithms for ontology matching. The Ontologies Mapping project is currently de ning the user requirements for a sustainable service, which will use such OM tools or algorithms. We will conduct a survey of Pistoia Alliance members to understand the need for such a service and whether it should be implemented in future.
This work was partially funded by the Pistoia Alliance Ontology Mappings project, the EU project Optique (FP7-ICT-318338), and the EPSRC projects ED3 and DBOnto.