Building a molecular glyco-phenotype ontology to decipher undiagnosed diseases

Hundreds of rare diseases are due to mutation on genes related to glycans synthesis, degradation or recognition. These glycan-related defects are well described in the literature but largely absent in ontologies and databases of chemical entities and phenotypes, limiting the application of computational methods and ontology-driven tools for characterization and discovery of glycan related diseases. We are curating articles and textbooks in glycobiology related to genetic diseases to inform the content and the structure of an ontology of Molecular Glyco-Phenotypes (MGPO). MGPO will be applied toward use cases including disease diagnosis and disease gene candidate prioritization, using semantic similarity and pattern matching at the glycan level with glycomics data from patient of the Undiagnosed Diseases Network.