Embedded Files
Date: April 14, 2017
Speaker: Clément Goehrs
Abstract
Drug-related data analysis is of public health interest. However, it remains complex. One of the main problematic is the multiplicity and heterogeneity of data sources both in France and internationally [1]. Drug-drug interaction analysis or drug misuse detection on social media are both good illustrations of this concern. There is a therefore a need for interoperability. Formal models, like ontologies, are one of the methods used to address this problematic. The objective of our work was to research and, if needed, develop an ontology that met our needs. We carried out a thorough bibliographic search that allowed selecting notably: DRON, CHEBI and DIDEO ontologies, the VANDF and NDFRT and RxNorm [2] models. These models all have interesting concepts, definitions and characteristics. Based on them, we developed a semantic drug model aiming to a modular architecture, for easy mapping with other ontologies. We then developed the ontology based on this formal model. Data from the French National Drug Database were instantiated, initially in order to test its consistency. This resource could be very useful in a number of cases related to French pharmacology, such as drugs misuse detection.
References
- Jovanovik M, Trajanov D. Consolidating drug data on a global scale using Linked Data. Journal of Biomedical Semantics. 2017 Jan 21;8(1):3.
- Nelson SJ, Zeng K, Kilbourne J, Powell T, Moore R. Normalized names for clinical drugs: RxNorm at 6 years. Journal of the American Medical Informatics Association. 2011 Jul 1;18(4):441-8.
Google Sites
Report abuse