Improving disease model accessibility at mouse genome informatics: making the move from OMIM to the disease ontology
Year: 2017
Type: other
Abstract: Mouse Genome Informatics (MGI, www.informatics.jax.org) has been using OMIM terms for annotations of mouse models to human disease for many years. While the OMIM disease records contain user friendly text descriptions of diseases and the OMIM group curates relationships between disease and human genes, the lack of disease groupings and relationships among diseases creates issues for display, retrieval and computational analysis of disease data. MGI has been working with the Rat Genome Database (RGD) and the Disease Ontology (DO) groups to enhance the DO. This work has focused on increasing and refining the incorporation of OMIM terms in the DO and enriching the relationships between disease terms in the DO. As a result of this work the DO has reached a level of development where MGI has begun to convert from OMIM to DO for display and searching for disease related data. The first phase of this conversion focused on the public web interface of MGI. Data annotations made to OMIM terms are translated into DO annotations using the OMIM xrefs in the DO file. The translated annotations are then used to populate disease related pages and searches in MGI. For example the disease tables shown on the marker detail pages now show DO terms and IDs. The disease browser and detail pages have been completely reworked to take advantage of the DAG structure of the DO. A user may now see all mouse models for Parkinson Disease, including all of the subtypes, on a single page rather than having to visit each subtype page individually. The Human – Mouse: Disease Connection has also been updated to use the DO. Users may search by DO terms names, synonyms, or IDs and may also search by many of the alternate IDs mapped to DO terms including OMIM, MESH, NCI, KEGG, ICD10, ICD9 and UMLS. In addition, the HMDC gene homologs x phenotype/disease page now groups diseases by a custom MGI DO slim set to improve visualization. By using the OMIM xrefs in the DO file to translate annotations we are able to continue to work on the coverage and representation of OMIM terms in DO without losing data due to incomplete coverage of OMIM or having to re-curate data following changes to OMIM representation in the DO. Future changes include annotating directly to DO terms and providing computational access to these data through APIs.
Authors: Susan M. Bello, Janan Eppig, Cynthia Smith
Institution: Jackson Laboratory
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: gold
OpenAlex ID: W2728068661