Sindhu, M S; Dr.Kannan, B(Cochin University of Science And Technology, July 23, 2013)
[+]
[-]
Abstract:
In the current study, epidemiology study is done by means of
literature survey in groups identified to be at higher potential for DDIs as
well as in other cases to explore patterns of DDIs and the factors affecting
them. The structure of the FDA Adverse Event Reporting System (FAERS)
database is studied and analyzed in detail to identify issues and challenges
in data mining the drug-drug interactions. The necessary pre-processing
algorithms are developed based on the analysis and the Apriori algorithm is
modified to suit the process. Finally, the modules are integrated into a tool to identify DDIs. The results are compared using standard drug interaction
database for validation. 31% of the associations obtained were identified to
be new and the match with existing interactions was 69%. This match
clearly indicates the validity of the methodology and its applicability to
similar databases. Formulation of the results using the generic names
expanded the relevance of the results to a global scale. The global
applicability helps the health care professionals worldwide to observe
caution during various stages of drug administration thus considerably
enhancing pharmacovigilance
Description:
Department of Computer Applications, Cochin
University of Science and Technology