The Challenge of Predicting Drug Toxicity in silico

Abstract

Poor pharmacokinetics, side effects and compound toxicity are frequent causes of late-stage failures in drug development. A safe in silico identification of adverse effects triggered by drugs and chemicals would be highly desirable as it not only bears economical potential but also spawns a variety of ecological benefits: sustainable resource management, reduction of animal models and possibly less risky clinical trials. In computer-aided drug discovery, both existing and hypothetical compounds may be studied; the methods are fast, reproducible, and typically based on human bioregulators, making the question of transferability obsolete. In the recent past, our laboratory contributed towards the development of in silico concepts (» multi-dimensional QSAR) and validated a series of ''virtual test kits'' based on the oestrogen, androgen, thyroid, and aryl hydrocarbon receptor (endocrine disruption, receptor-mediated toxicity) as well as on the enzyme cytochrome P450 3A4 (metabolic transformations, drug-drug interactions). The test kits are based on the threedimensional structure of their target protein (i.e. ER ab , AR, TR ab , CYP450) or a surrogate thereof (AhR) and were trained using a representative selection of 362 substances. Subsequent evaluation of 107 compounds different therefrom showed that binding affinities are predicted close to experimental uncertainty. These results suggest that our approach is suited for the in silico identification of adverse effects triggered by drugs and chemicals and encouraged us to compile an Internet Database for the virtual screening of drugs and chemicals for toxic effects.

VirtualToxLab ~ in silica prediction of the toxic potential of drugs and environmental chemicals: evaluation status and Internet access protocol

“…Vedani et al, 2005b). m...... (Vedani et al, 2006); the Raptor model depicts the thyroid [3 receptor (Vedani at al., 2007).…”

“…This protocol was applied to the estrogen receptor alß. (Vedani et al, 2005b;Peristera and Vedani, in preparation), androgen receptor , the peroxisome proliferator-activated receptor y (Vedani et al , 2007), the glucocorticoid receptor (Spreafico et al, in preparation) and the enzyme cytochrome P450 3A4 (Lill et al, 2006). For the aryl hydrocarbon receptor, for which no experimental structure is available, the alignment was performed based on the molecular skeletons (Vedani et al, 1999a(Vedani et al, , 1999b(Vedani et al, , 2003(Vedani et al, , 2005aVedani and Dobler, 2002).…”

“…the 154 aryl hydrocarbon receptor). In this context, we have suggested the term "toxic potential", as the binding to a key protein is aprerequisite for triggering an adverse effect (Vedani et al, 2005a(Vedani et al, , 2006.…”

In Silico-In Vitro Screening of Protein-Protein Interactions: Towards the Next Generation of Therapeutics

“…As introduced above, many purchasable compounds or virtual compounds may not have satisfactory ADME/Tox properties to be considered as promising drug candidates. It is then possible to remove from the libraries some of these molecules by computing properties such as: absorption prediction via computation of logP, logD, molecular weight, polar surface area (e.g., if greater than 140 Å 2 , poor intestinal absorption), number of H-bond donors-acceptors, number of rotatable bonds (see Lipinski rule of 5 reviewed in [191]), frequent hitters, reactive chemical groups, etc) using in silico methods ( [191][192][193][194][195][196][197][198][199]) (Table III). It is commonly accepted that absorption, distribution and excretion are dependent on similar descriptors while metabolism and toxicity are of a quite different nature and depend of numerous factors.…”

Exploring the Ligand-Protein Networks in Traditional Chinese Medicine: Current Databases, Methods and Applications

“…However, in the recent years, more and more evidence has shown that many drugs exert their activities by modulating multitargets [2][3][4]. Besides, some drugs interact with antitargets and induce strong side effects [5,6]. Therefore, it is inappropriate to stick to the paradigm that drug interacts with only one target.…”