Ensemble Models Based on QuBiLS-MAS Features and Shallow Learning for the Prediction of Drug-Induced Liver Toxicity: Improving Deep Learning and Traditional Approaches

Abstract

Drug-induced liver injury (DILI) is a key safety issue in the drug discovery pipeline and a regulatory concern. Thus, many in silico tools have been proposed to improve the hepatotoxicity prediction of organic-type chemicals. Here, classifiers for the prediction of DILI were developed by using QuBiLS-MAS 0−2.5D molecular descriptors and shallow machine learning techniques, on a training set composed of 1075 molecules. The best ensemble model build, E13, was obtained with good statistical parameters for the learning series, namely, the following: accuracy = 0.840, sensibility = 0.890, specificity = 0.761, Matthew's correlation coefficient = 0.660, and area under the ROC curve = 0.904. The model was also satisfactorily evaluated with Y-scrambling test, and repeated k-fold cross-validation and repeated k-holdout validation. In addition, an exhaustive external validation was also carried out by using two test sets and five external test sets, with an average accuracy value equal to 0.854 (±0.062) and a coverage equal to 98.4% according to its applicability domain. A statistical comparison of the performance of the E13 model, with regard to results and tools (e.g., Padel DDPredictor Software, Deep Learning DILIserver, and Vslead) reported in the literature, was also performed. In general, E13 presented the best global performance in all experiments. The sum of the ranking differences procedure provided a very similar grouping pattern to that of the M-ANOVA statistical analysis, where E13 was identified as the best model for DILI predictions. A noncommercial and fully cross-platform software for the DILI prediction was also developed, which is freely available at http://tomocomd.com/apps/ptoxra. This software was used for the screening of seven data sets, containing natural products, leads, toxic materials, and FDA approved drugs, to assess the usefulness of the QSAR models in the DILI labeling of organic substances; it was found that 50−92% of the evaluated molecules are positive-DILI compounds. All in all, it can be stated that the E13 model is a relevant method for the prediction of DILI risk in humans, as it shows the best results among all of the methods analyzed.

In Silico Searching for Alternative Lead Compounds to Treat Type 2 Diabetes through a QSAR and Molecular Dynamics Study

“…The applicability domain analysis used in this study is based on a consensus score of four methods. The score represents the fraction of the four methods that indicate a molecule is considered inside the applicability domain, with scores ranging from zero (molecule is identified as an outlier by all four methods) to one (molecule is inside the AD for all four methods) [21,62]. A score greater than 0.25 was used as the criteria for identifying compounds inside the AD.…”

“…In silico analysis is an economic strategy for the exploration of new compounds [21,22]. Quantitative structure-activity relationship (QSAR) reveals relationships between structural descriptors and biological activities of chemical compounds.…”

Toxicity prediction based on artificial intelligence: A multidisciplinary overview

“…More recently, Minerali et al 58 have generated and compared ML algorithms for predicting DILI with in‐house software Assay Central (https://www.collaborationspharma.com/assay-central), using data previously collated by Pfizer and AstraZeneca research groups, along with data from the FDA, performing an ROC of 81%, sensitivity of 74%, specificity of 76%, and accuracy of 75% for the best Bayesian model based on the DILI‐concern category from the DILI Rank database. Noncommercial software has been developed by Mora et al, freely available at http://tomocomd.com/apps/ptoxra for the DILI prediction with ML models on a training set of 1075 molecules, attaining an accuracy of 84%, a sensibility of 89%, specificity of 76%, and ROC of 90%; 59 In the pharmaceutical company AstraZeneca Williams et al 60 predicted DILI by using ML, quantifying the probability that a compound falls into either low, medium, or high‐severity categories, with an accuracy of 63%. For a binary yes/no DILI prediction, the model obtained an accuracy of 86%, sensitivity of 87%, specificity of 85%, a positive predictive value of 92%, and a negative predictive value of 78%.…”

Molecular insights on ABL kinase activation using tree-based machine learning models and molecular docking

“…Well-known examples of these models are the decision tree (DT) [ 55 ], random forest (RF) [ 56 ], and adaptive boosting (AdaBoost) [ 57 ]. Tree-based models are widely used in several research areas, such as metabolomics [ 58 ], disease detection [ 59 ], toxicological predictions [ 60 – 62 ], and stock prediction [ 63 ], due to their simplicity, efficiency, and interpretability.…”