QSAR Modeling and Data Mining Link Torsades de Pointes Risk to the Interplay of Extent of Metabolism, Active Transport, and hERG Liability
We collected 1173 hERG patch clamp (PC) data (IC(50)) from the literature to derive twelve classification models for hERG inhibition, covering a large variety of chemical descriptors and classification algorithms. Models were generated using 545 molecules and validated through 258 external molecules tested in PC experiments. We also evaluated the suitability of the best models to predict the activity of 26 proprietary compounds tested in radioligand binding displacement (RBD). Results proved the necessity to use multiple validation sets for a true estimation of model accuracy and demonstrated that using various descriptors and algorithms improves the performance of ligand-based models. Intriguingly, one of the most accurate models uncovered an unexpected link between extent of metabolism and hERG liability. This hypothesis was fairly reinforced by using the Biopharmaceutics Drug Disposition Classification System (BDDCS) that recognized 94% of the hERG inhibitors as extensively metabolized in vivo. Data mining suggested that high Torsades de Pointes (TdP) risk results from an interplay of hERG inhibition, extent of metabolism, active transport, and possibly solubility. Overall, these new findings might improve both the decision making skills of pharmaceutical scientists to mitigate hERG liability during the drug discovery process and the TdP risk assessment during drug development.
[Show abstract] [Hide abstract] ABSTRACT: The Rule of 5 methodology appears to be as useful today in defining drugability as when it was proposed, but recognizing that the database that we used includes only drugs that successfully reached the market. We do not view additional criteria necessary nor did we find significant deficiencies in the four Rule of 5 criteria originally proposed by Lipinski and coworkers. BDDCS builds upon the Rule of 5 and can quite successfully predict drug disposition characteristics for drugs both meeting and not meeting Rule of 5 criteria. More recent expansions of classification systems have been proposed and do provide useful qualitative and quantitative predictions for clearance relationships. However, the broad range of applicability of BDDCS beyond just clearance predictions gives a great deal of further usefulness for the combined Rule of 5/BDDCS system.0Comments 0Citations
- "BDDCS extensions have begun to appear related to toxicity predictions and environmental implications. Broccatelli et al.  used the BDDCS to help consider for which drugs hERG voltage-gated potassium channel inhibition is likely to lead to Torsade de Pointes. Vuppalanchi et al.  and we  utilized BDDCS in evaluating drug induced liver injury, while our laboratory has also proposed the use of BDDCS to predict which anti-epileptic drugs will cause drug hypersensitivity reactions . "
- "Most of the previously available hERG computational models are ligand-based but since these rely mainly on the existence of structural similarities between the tested drug candidate and previously reported known hERG blockers, these models have limited value in the frequent case of a novel drug structure (Aronov and Goldman, 2004; Coi et al., 2006; Du-Cuny et al., 2011; Ekins et al., 2002; Keseru, 2003; Song and Clark, 2006; Su et al., 2010; Yoshida and Niwa, 2006). Producing a reliable hERG ion channel structure-based model to directly test drug binding (Broccatelli et al., 2012; Du-Cuny et al., 2011) has therefore been of high priority but in the absence of a hERG crystal structure, this has been limited to creating homology models that have excluded a large portion of the hERG channel and which display very limited conformational flexibility (Boukharta et al., 2011; Di Martino et al., 2013; Farid et al., 2006; Osterberg and Aqvist, 2005). Previous models have also relied mainly on docking scoring functions to predict binding energies while ignoring the exact binding location of the tested ligands relative to the mouth of the ion channel. "
[Show abstract] [Hide abstract] ABSTRACT: Most computational methods used for the prediction of toxicity endpoints are based on the assumption that similar compounds have similar biological properties. This principle can be exploited using computational methods like read across or quantitative structure-activity relationships. However, there is no general agreement about which method is the most appropriate for quantifying compound similarity neither for exploiting the similarity principle in order to obtain reliable estimations of the compound properties. Moreover, optimal similarity metrics and modeling methods might depend on the characteristics of the endpoints and training series used in each case. This study describes a comparative analysis of the predictive performance of diverse similarity metrics and modeling methods in toxicological applications. A collection of two quantitative (n = 660, n = 1114) and three qualitative (n = 447, n = 905, n = 1220) datasets representing very different endpoints of interest in drug safety evaluation and rigorous methods were used to estimate the external predictive ability in each case. The results confirm that no single approach produces the best results in all instances, and the best predictions were obtained using different tools in different situations. The trends observed in this study were exploited to propose a unifying strategy allowing the use of the most suitable method for every compound. A comparison of the quality of the predictions obtained by the unifying strategy with those obtained by standard prediction methods confirmed the usefulness of the proposed approach.0Comments 1Citation
- "The development of such models is further hampered by the low quality of training data extracted from bibliographic sources, due to the large disparity of the experimental conditions (e.g., cell line, temperature, pH) used in the original experiments. The ABCB1 dataset (Broccatelli et al. 2012) contains 562 inhibitors and 515 non-inhibitors of ABCB1. ABCB1, also known as P-glycoprotein, is a membrane protein member of the ATP-binding cassette (ABC) transporters superfamily , which transports a variety of compounds through the membrane against a concentration gradient (Choudhuri and Klaassen 2006). "