The In Silico Prediction of Human-Specific Metabolites from Hepatotoxic Drugs

Ravouru      Nagaraju; Ande   Penchala   Prathusha; Rajesh      Kaza; Koganti      Bharathi; Luis      G. Valerio; Anthony      Long

Abstract

In this study we employed the use of the Meteor computational software program to perform predictions in silico on 17 hepatotoxic drugs for determining human-specific drug metabolites. Congruence of the in silico predictions from a qualitative standpoint of drug metabolite structures was established by comparison to human in vivo drug metabolic profiles characterized in publically available clinical studies. A total of 87 human-specific metabolites were identified from the 17 drugs. We found that Meteors positive predictions included 4 out of the 9 reported major metabolites (detected in excreta at a level of > 10% of the administered p.o. dose) and 10 out of the 15 major phase II metabolites giving a total of 14 correctly predicted drug metabolite structures out of 23 major metabolites. A significant level of unconfirmed positive predictions resulted and discussion on reasons for this is presented. An example is given whereby the in silico metabolism prediction succeeded to predict the putative toxic pathway of one of the drugs whilst conventional rodent liver microsomal assays failed to predict the pathway. Overall, we describe a reasonable simulation of human metabolic profiling using this in silico method with this data set of hepatotoxic drugs now withdrawn from the market. We provide an in-depth and objective discussion of this first of its kind validation test using clinical study data with interest in the prediction human-specific metabolism. Further research is discussed on what areas need to be investigated to improve upon the predictive data. The strong potential of this method to predict human-specific drug metabolites suggests the utility of this computational tool to help support not only the discovery development of therapeutics but also the safety assessment in identifying drug metabolites early to protect patients prior to initiating clinical studies.

Keywords: Computational drug metabolism, drug metabolism, in silico method, alternative methods, hepatotoxic drugs, liver microsomes, primary hepatocytes, liver slices, ADME-Tox, MedDRA, Drug Metabolites, Meteor, Biotransformations, trovafloxacin, flosequinan, lipophilicity, Beclobrate, Benoxaprofen, Dilevalol, Flosequinan Total, Glafenine, Indoprofen, Moxisylyte, Nimesulide, Niridazole, Nomifensine, Pirprofen, Practolol, Tasosartan, Ximelagatran, Zileuton, Zomepirac, parabenzoquinone, Multiple Biotransformation Pathways, HPLC, adverse drug re-actions, ADRs