Modeling Drug Mechanism Knowledge Using Evidence and Truth Maintenance

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.

[1]  J. S. Wang,et al.  Gemfibrozil is a potent inhibitor of human cytochrome P450 2C9. , 2001, Drug metabolism and disposition: the biological fate of chemicals.

[2]  Daniel Kahneman,et al.  Probabilistic reasoning , 1993 .

[3]  Janette M Carpenter,et al.  The Top 100 Drug Interactions: A Guide to Patient Management , 2002 .

[4]  Fumiyoshi Yamashita,et al.  In silico approaches for predicting ADME properties of drugs. , 2004, Drug metabolism and pharmacokinetics.

[5]  Ira J. Kalet,et al.  Qualitative Pharmacokinetic Modeling of Drugs , 2005, AMIA.

[6]  J. Feely,et al.  Pharmacokinetic-Pharmacodynamic Drug Interactions with HMG-CoA Reductase Inhibitors , 2002, Clinical pharmacokinetics.

[7]  Hartmut Derendorf,et al.  Pharmacokinetic/Pharmacodynamic Modeling in Drug Research and Development , 2000, Journal of clinical pharmacology.

[8]  P. Hansten,et al.  Drug interaction management , 2003, Pharmacy World and Science.

[9]  I. Kapetanovic,et al.  Stable isotope methodology and gas chromatography mass spectrometry in a pharmacokinetic study of phenobarbital. , 1980, Biomedical mass spectrometry.

[10]  Kenneth D. Forbus,et al.  Building Problem Solvers , 1993 .

[11]  H. van de Waterbeemd,et al.  ADMET in silico modelling: towards prediction paradise? , 2003, Nature reviews. Drug discovery.

[12]  J. S. Wang,et al.  In vitro evaluation of valproic acid as an inhibitor of human cytochrome P450 isoforms: preferential inhibition of cytochrome P450 2C9 (CYP2C9). , 2001, British journal of clinical pharmacology.

[13]  Judea Pearl,et al.  Qualitative Probabilities for Default Reasoning, Belief Revision, and Causal Modeling , 1996, Artif. Intell..

[14]  J. Houston,et al.  In vitro cytochrome P450 inhibition data and the prediction of drug‐drug interactions: Qualitative relationships, quantitative predictions, and the rank‐order approach , 2005, Clinical pharmacology and therapeutics.

[15]  Kiyomi Ito,et al.  Database analyses for the prediction of in vivo drug-drug interactions from in vitro data. , 2004, British journal of clinical pharmacology.

[16]  P. Bonnabry,et al.  Quantitative Drug Interactions Prediction System (Q-DIPS) , 2001, Clinical pharmacokinetics.

[17]  J P Rindone,et al.  Gemfibrozil-warfarin drug interaction resulting in profound hypoprothrombinemia. , 1998, Chest.

[18]  Judea Pearl,et al.  Bayesian Networks , 1998, Encyclopedia of Social Network Analysis and Mining. 2nd Ed..

[19]  Simon Parsons,et al.  A review of uncertainty handling formalisms , 1998, Applications of Uncertainty Formalisms.

[20]  Peter Norvig,et al.  Artificial Intelligence: A Modern Approach , 1995 .

[21]  Edward H. Shortliffe,et al.  Production Rules as a Representation for a Knowledge-Based Consultation Program , 1977, Artif. Intell..

[22]  W. Theodore,et al.  Mechanism of valproate‐phenobarbital interaction in epileptic patients , 1981, Clinical pharmacology and therapeutics.

[23]  J. Lin,et al.  Sense and nonsense in the prediction of drug-drug interactions. , 2000, Current drug metabolism.

[24]  Edward H. Shortliffe,et al.  Production rules as a basis for a knowledge-based 867 consultation program , 1977 .

[25]  L J Lesko,et al.  Drug Interaction Studies: Study Design, Data Analysis, and Implications for Dosing and Labeling , 2007, Clinical pharmacology and therapeutics.

[26]  Peter D. Karp,et al.  An Evidence Ontology for Use in Pathway/Genome Databases , 2003, Pacific Symposium on Biocomputing.