Metabolic drug interactions are a major source of clinical problems, but their investigation during drug development is often incomplete and poorly specific. In vitro studies give very accurate data on the interactions of drugs with selective cytochrome P450 (CYP) isozymes, but their interpretation in the clinical context is difficult. On the other hand, the design of in vivo studies is sometimes poor (choice of prototype substrate, doses, schedule of administration, number of volunteers), with the risk of minimising the real potential for interaction. To link in vitro and in vivo studies, several authors have suggested using extrapolation techniques, based on the comparison of in vitro inhibition data with the active in vivo concentrations of the inhibitor. However, the lack of knowledge of one or several important parameters (role of metabolites, intrahepatocyte accumulation) often limits the possibility for safe and accurate predictions. In consequence, these methods are useful to complement in vitro studies and help design clinically relevant in vivo studies, but they will not totally replace in vivo investigation in the future. We have developed a computerised application, the quantitative drug interactions prediction system (Q-DIPS), to make both qualitative deductions and quantitative predictions on the basis of a database containing updated information on CYP substrates, inhibitors and inducers, as well as pharmacokinetic parameters. We also propose a global approach to drug interactions problems--'good interactions practice--to help design rational drug interaction investigations, sequentially associating in vitro studies, in vitrolin vivo extrapolation and finally well-designed in vivo clinical studies.
"While other drug knowledge-bases and knowledge-based systems have linked evidence to their drug facts (e.g. DRUGDEX Ò6 , Q-DIPS , PharmGKB , and BioCyc ), only the DIKB classifies all evidence entered into the system using a biomedical evidence ontology oriented toward confidence assignment. This enables the system to provide customized views of a body of drug-mechanism knowledge to users who do not agree about the inferential value of particular evidence types. "
[Show abstract][Hide abstract] ABSTRACT: We describe a novel experiment that we conducted with the Drug Interaction Knowledge-base (DIKB) to determine which combinations of evidence enable a rule-based theory of metabolic drug-drug interactions to make the most optimal set of predictions. The focus of the experiment was a group of 16 drugs including six members of the HMG-CoA-reductase inhibitor family (statins). The experiment helped identify evidence-use strategies that enabled the DIKB to predict significantly more interactions present in a validation set than the most rigorous strategy developed by drug experts with no loss of accuracy. The best-performing strategies included evidence types that would normally be of lesser predictive value but that are often more accessible than more rigorous types. Our experimental methods represent a new approach to leveraging the available scientific evidence within a domain where important evidence is often missing or of questionable value for supporting important assertions.
"Interestingly, while labeling is the primary source of drug mechanism knowledge for DRUGDEX ® , Facts and Comparisons ® , and other comparable systems, these systems use editorial boards to stay current with knowledge from clinical trials or case reports and it has proven scalable to the thousands of drug products listed in these sources. Q-DIPS , a system designed to help clinicians identify and manage DDIs that occur by metabolic mechanisms, demonstrates that the editorial board approach is feasible for drug-mechanism knowledge. The system's maintainers curate a database of in vitro and in vivo studies that support assertions about the enzymes a drug is a substrate of or modulates. "
[Show abstract][Hide abstract] ABSTRACT: To protect the safety of patients, it is vital that researchers find methods for representing drug mechanism knowledge that support making clinically relevant drug-drug interaction (DDI) predictions. Our research aims to identify the challenges of representing and reasoning with drug mechanism knowledge and to evaluate potential informatics solutions to these challenges through the process of developing a knowledge-based system capable of predicting clinically relevant DDIs that occur via metabolic mechanisms. In previous work, we designed a simple, rule-based, model of metabolic inhibition and induction and applied it to a database containing assertions about 267 drugs. This pilot system taught us that drug mechanism knowledge is often dynamic, missing, or uncertain. In this paper, we propose methods to address these properties of mechanism knowledge and describe a new prototype system, the Drug Interaction Knowledge-base (DIKB), that implements our proposed methods so that we can explore their strengths and limitations. A novel feature of the DIKB is its use of a truth maintenance system to link changes in the evidence support for assertions about drug properties to the set of interactions and non-interactions the system predicts.
IEEE Transactions on Information Technology in Biomedicine 08/2007; 11(4):386-97. DOI:10.1109/TITB.2007.890842 · 2.49 Impact Factor
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.