Skip to Main content Skip to Navigation
Master Thesis

Interactions médicamenteuses : données liées et applications

Abstract : Introduction. One significant cause of adverse drug reactions is drug-drug interaction (DDI). In France, the thesaurus of DDI edited by ANSM (french national drug safety insti- tute) is the main referential. However, its PDF format impedes its utilisation. The objective of this work was to propose an open format of the thesaurus and illustrate its usefulness for research through two examples : to examine the overlap between the thesaurus and other sources and to estimate the prevalence of potential DDIs based on drug deliveries. Methods. We developped a program that transforms the PDF format to RDF triples, a web semantic standard for data exchange. We linked its content to three sources : the Unified Medical Language System (UMLS), DrugBank (DB) and ’le répertoire des médicaments’, a french drug database edited by ANSM. The link to DB enables its integration into an international DDI-database (merged-PDDI) and the measure of overlap with other sources. A program was created to estimate the prevalence of potential DDIs based on the thesaurus open format and drug deliveries data. Results. In its version of january 2016, the french thesaurus contains 1102 drugs and 168 drug classes. It describes 1414 interactions between classes or molecules resulting in 52870 DDI. Among the latters, 5651 (13%) were found in the merged-PDDI dataset corresponding to little overlap with other sources. We identified 17112 contraindicated drug dispensations among 7 millions drug deliveries between june and august 2013. Conclusion. We propose a RDF format of the french thesaurus of DDI linked to UMLS, DB and the ’repertoire des médicaments’. This format can facilitate research on DDI and the development of new applications.
Document type :
Master Thesis
Complete list of metadatas

https://dumas.ccsd.cnrs.fr/dumas-01442668
Contributor : Bu Carreire Université de Bordeaux <>
Submitted on : Friday, January 20, 2017 - 6:25:26 PM
Last modification on : Friday, September 13, 2019 - 5:26:02 PM
Long-term archiving on: : Friday, April 21, 2017 - 4:34:08 PM

Identifiers

  • HAL Id : dumas-01442668, version 1

Collections

Citation

Sébastien Cossin. Interactions médicamenteuses : données liées et applications. Médecine humaine et pathologie. 2016. ⟨dumas-01442668⟩

Share

Metrics

Record views

222

Files downloads

2466