FBOnto: An ontology to represent food intake data and associate it with metabolomic data

Currently, nutrition research generates a lot of complex data hard to analyze and associate with other data or omics such as metabolomics. By itself, metabolomics is closely linked to nutrition. However, it is still difficult to associate these two types of data. One reason for this difficulty is the heterogeneity found in the information provided by participants in nutritional studies about what they have eaten. In order to manage this heterogeneity we have decide to build an ontology, describing foods in a hierarchichal way that enables a common description of food intake. This ontology contains formal naming, definition of the categories, properties, and relations between the concepts of two types of data, food and related metabolites.

This ontology has been created using Protégé 5.2.0 version. As is common in ontologies, FBOnto (Food-Biomarker Ontology) has been written in OWL language (Web Ontology Language).

The ontology presented is called FBOnto and is made of two sub-ontologies. The first ontology describes foods: from simple such as fruits and vegetables to more complex foods such as ratatouille. The second ontology describes metabolites, grouped in their different chemical classes. The nodes or elements of these two sub-ontologies are connected by the properties of each one, so that if a metabolite is in different foods, it will connect with all of them. This ontology allows us to visualize data in a bidirectional way, going from metabolomics to nutritional data or vice versa. In addition, this ontology can also be useful for other analyses such as enrichment analysis or a novel concept, food-enrichment analysis.