Chinese-English Ontology Construction And Knowledge Discovery Of Adverse Events Associated With Cardiovascular Drugs Used In China

Objectives:To construct the Chinese-English ontology knowledge base and systematically represent adverse events(AEs)associated with cardiovascular drugs used in China;to provide resources and foundation for systematic analysis and knowledge discovery of cardiovascular drug adverse events;To analyse the pattern of cardiovascular drug adverse events based on the ontology;To develop the data mining methods for the discovery of the drug class effects on adverse drug events.Methods:The cardiovascular drugs(CVDs)used in China according to New Pharmacology(17th version)were collected,and the cardiovascular drug package inserts were collected.The drug names and AE terms from drug package inserts were extracted and translated.AE terms were mapped to OAE(the Ontology of Adverse Events)terms.If there was no mapping,the new term was annotated and added into OAE.Drug names were mapped to NDF-RT(The National Drug File-Reference Terminology).OCVDAE(The Ontology of Cardiovascular Drug Adverse Events)was constructed through Onto Fox and Ontorat.The ingredient classification and mechanisms of action(Mo A)classification in NDF-RT were integrated into the OCVDAE,and the CVD-AE associations were added into OCVDAE.Based on the ontology,the pattern of cardiovascular drug adverse events was statistically analyzed,including the drug classes,adverse drug event classes and top adverse drug events.Using data mining methods,including PCR(drug proportional class level ratio),C-PRR(the drug class level proportional reporting ratio),C-χ2(the drug class-χ2),minimal number of drugs in a drug class and heatmap,based on the AE classification,the drug ingredient classification and Mo A classification in OCVDAE,the study mined statistically significant drug class effect on adverse drug events.Finally,some of the drug class effects knowledge was confirmed through the literatures in Pub Med.Results:265 cardiovascular drugs and the corresponding drug package inserts were collected,1383 Chinese AE terms and associated with 802 English terms were extracted.Among these 802 terms,391 were mapped to OAE,411 terms were annotated and added to OAE and formed new OAE terms.In total,194 drugs were mapped to NDF-RT.The OCVDAE can access openly on http://www.ontobee.org/ontology/and http://purl.bioontology.org/ontology/OCVDAE.The OCVDAE included 194 drugs and 736 AE terms.265 CVDs are associated with 9 drug classes,802 AE terms are associated with 27 AE classes and 10 top AEs include headache,nausea,fatigue,etc.PCR-based heatmap analysis identified many class level drug effects on different AE classes such as behavioral and neurological AE and digestive system AE.The ingredient class lipids(including atorvastatin,fluvastatin,and fenofibrate)was found to have a statistically significant class effect on 8 adverse drug events(e.g.,insomnia,myopathy,myalgia,and myositis AEs).Three Mo A classes(e.g.,Cholesterol Synthesis Inhibitors)have statistically significant class effects on 5 adverse drug events(e.g.,anorexia,diarrhea AEs).Some results of these methods were confirmed by literatures.Conclusion:This study systematically collected cardiovascular drugs used in China and CVD-associated AEs from package inserts,generated the OCVDAE ontology by reusing terms from the OAE and NDF-RT and adding CVD-AE associations.Based on the ontology,the pattern of cardiovascular drug adverse events was statistically analysed.PCR scores,C-PRR,C-χ2,minimal number of drugs in a drug class for a specific AE(or AE class)were applied to find the drug class effect.