Development and evaluation of in silico prediction model for drug-induced respiratory toxicity by using naïve Bayes classifier method

Improvement of a prediction model for heart failure survival through explainable artificial intelligence

Cardiovascular diseases and their associated disorder of heart failure (HF) are major causes of death globally, making it a priority for doctors to detect and predict their onset and medical consequences. Artificial Intelligence (AI) allows doctors to discover clinical indicators and enhance their diagnoses and treatments. Specifically, "eXplainable AI" (XAI) offers tools to improve the clinical prediction models that experience poor interpretability of their results. This work presents an explainability analysis and evaluation of two HF survival prediction models using a dataset that includes 299 patients who have experienced HF. The first model utilizes survival analysis, considering death events and time as target features, while the second model approaches the problem as a classification task to predict death. The model employs an optimization data workflow pipeline capable of selecting the best machine learning algorithm as well as the optimal collection of features. Moreover, different post hoc techniques have been used for the explainability analysis of the model. The main contribution of this paper is an explainability-driven approach to select the best HF survival prediction model balancing prediction performance and explainability. Therefore, the most balanced explainable prediction models are Survival Gradient Boosting model for the survival analysis and Random Forest for the classification approach with a c-index of 0.714 and balanced accuracy of 0.74 (std 0.03) respectively. The selection of features by the SCI-XAI in the two models is similar where "serum_creatinine", "ejection_fraction", and "sex" are selected in both approaches, with the addition of "diabetes" for the survival analysis model. Moreover, the application of post hoc XAI techniques also confirm common findings from both approaches by placing the "serum_creatinine" as the most relevant feature for the predicted outcome, followed by "ejection_fraction". The explainable prediction models for HF survival presented in this paper would improve the further adoption of clinical prediction models by providing doctors with insights to better understand the reasoning behind usually "black-box" AI clinical solutions and make more reasonable and data-driven decisions.

Development of binary classification models for assessment of drug-induced liver injury in humans using a large set of FDA-approved drugs

Drug-induced liver injury (DILI) has been identified as one of the major causes for drugs withdrawn from the market, and even termination during the late stages of development. Therefore, it is imperative to evaluate the DILI potential of lead compounds during the research and development process. Although various computational models have been developed to predict DILI, most of which applied the DILI data were extracted from preclinical sources. In this investigation, the in silico prediction models for DILI were constructed based on 1140 FDA-approved drugs by using naïve Bayes classifier approach. The genetic algorithm method was applied for the molecular descriptors selection. Among these established prediction models, the NB-11 model based on eight molecular descriptors combined with ECFP_18 showed the best prediction performance for DILI, which gave 91.7% overall prediction accuracy for the training set, and 68.9% concordance for the external test set. Therefore, the established NB-11 prediction model can be used as a reliable virtual screening tool to predict DILI adverse effect in the early stages of drug design. In addition, some new structural alters for DILI were identified, which could be used for structural optimization in the future drug design by medicinal chemists.

An Explainable Supervised Machine Learning Model for Predicting Respiratory Toxicity of Chemicals Using Optimal Molecular Descriptors

Respiratory toxicity is a serious public health concern caused by the adverse effects of drugs or chemicals, so the pharmaceutical and chemical industries demand reliable and precise computational tools to assess the respiratory toxicity of compounds. The purpose of this study is to develop quantitative structure-activity relationship models for a large dataset of chemical compounds associated with respiratory system toxicity. First, several feature selection techniques are explored to find the optimal subset of molecular descriptors for efficient modeling. Then, eight different machine learning algorithms are utilized to construct respiratory toxicity prediction models. The support vector machine classifier outperforms all other optimized models in 10-fold cross-validation. Additionally, it outperforms the prior study by 2% in prediction accuracy and 4% in MCC. The best SVM model achieves a prediction accuracy of 86.2% and a MCC of 0.722 on the test set. The proposed SVM model predictions are explained using the SHapley Additive exPlanations approach, which prioritizes the relevance of key modeling descriptors influencing the prediction of respiratory toxicity. Thus, our proposed model would be incredibly beneficial in the early stages of drug development for predicting and understanding potential respiratory toxic compounds.

FP-ADMET: a compendium of fingerprint-based ADMET prediction models

MOTIVATION

The absorption, distribution, metabolism, excretion, and toxicity (ADMET) of drugs plays a key role in determining which among the potential candidates are to be prioritized. In silico approaches based on machine learning methods are becoming increasing popular, but are nonetheless limited by the availability of data. With a view to making both data and models available to the scientific community, we have developed FPADMET which is a repository of molecular fingerprint-based predictive models for ADMET properties. In this article, we have examined the efficacy of fingerprint-based machine learning models for a large number of ADMET-related properties. The predictive ability of a set of 20 different binary fingerprints (based on substructure keys, atom pairs, local path environments, as well as custom fingerprints such as all-shortest paths) for over 50 ADMET and ADMET-related endpoints have been evaluated as part of the study. We find that for a majority of the properties, fingerprint-based random forest models yield comparable or better performance compared with traditional 2D/3D molecular descriptors.

AVAILABILITY

The models are made available as part of open access software that can be downloaded from https://gitlab.com/vishsoft/fpadmet .

Machine learning models for classification tasks related to drug safety

In this review, we outline the current trends in the field of machine learning-driven classification studies related to ADME (absorption, distribution, metabolism and excretion) and toxicity endpoints from the past six years (2015-2021). The study focuses only on classification models with large datasets (i.e. more than a thousand compounds). A comprehensive literature search and meta-analysis was carried out for nine different targets: hERG-mediated cardiotoxicity, blood-brain barrier penetration, permeability glycoprotein (P-gp) substrate/inhibitor, cytochrome P450 enzyme family, acute oral toxicity, mutagenicity, carcinogenicity, respiratory toxicity and irritation/corrosion. The comparison of the best classification models was targeted to reveal the differences between machine learning algorithms and modeling types, endpoint-specific performances, dataset sizes and the different validation protocols. Based on the evaluation of the data, we can say that tree-based algorithms are (still) dominating the field, with consensus modeling being an increasing trend in drug safety predictions. Although one can already find classification models with great performances to hERG-mediated cardiotoxicity and the isoenzymes of the cytochrome P450 enzyme family, these targets are still central to ADMET-related research efforts.

Machine Learning in Predictive Toxicology: Recent Applications and Future Directions for Classification Models

In recent times, machine learning has become increasingly prominent in predictive toxicology as it has shifted from in vivo studies toward in silico studies. Currently, in vitro methods together with other computational methods such as quantitative structure-activity relationship modeling and absorption, distribution, metabolism, and excretion calculations are being used. An overview of machine learning and its applications in predictive toxicology is presented here, including support vector machines (SVMs), random forest (RF) and decision trees (DTs), neural networks, regression models, naïve Bayes, k-nearest neighbors, and ensemble learning. The recent successes of these machine learning methods in predictive toxicology are summarized, and a comparison of some models used in predictive toxicology is presented. In predictive toxicology, SVMs, RF, and DTs are the dominant machine learning methods due to the characteristics of the data available. Lastly, this review describes the current challenges facing the use of machine learning in predictive toxicology and offers insights into the possible areas of improvement in the field.

DeepDILI: Deep Learning-Powered Drug-Induced Liver Injury Prediction Using Model-Level Representation

Drug-induced liver injury (DILI) is the most frequently reported single cause of safety-related withdrawal of marketed drugs. It is essential to identify drugs with DILI potential at the early stages of drug development. In this study, we describe a deep learning-powered DILI (DeepDILI) prediction model created by combining model-level representation generated by conventional machine learning (ML) algorithms with a deep learning framework based on Mold2 descriptors. We conducted a comprehensive evaluation of the proposed DeepDILI model performance by posing several critical questions: (1) Could the DILI potential of newly approved drugs be predicted by accumulated knowledge of early approved ones? (2) is model-level representation more informative than molecule-based representation for DILI prediction? and (3) could improved model explainability be established? For question 1, we developed the DeepDILI model using drugs approved before 1997 to predict the DILI potential of those approved thereafter. As a result, the DeepDILI model outperformed the five conventional ML algorithms and two state-of-the-art ensemble methods with a Matthews correlation coefficient (MCC) value of 0.331. For question 2, we demonstrated that the DeepDILI model's performance was significantly improved (i.e., a MCC improvement of 25.86% in test set) compared with deep neural networks based on molecule-based representation. For question 3, we found 21 chemical descriptors that were enriched, suggesting a strong association with DILI outcome. Furthermore, we found that the DeepDILI model has more discrimination power to identify the DILI potential of drugs belonging to the World Health Organization therapeutic category of 'alimentary tract and metabolism'. Moreover, the DeepDILI model based on Mold2 descriptors outperformed the ones with Mol2vec and MACCS descriptors. Finally, the DeepDILI model was applied to the recent real-world problem of predicting any DILI concern for potential COVID-19 treatments from repositioning drug candidates. Altogether, this developed DeepDILI model could serve as a promising tool for screening for DILI risk of compounds in the preclinical setting, and the DeepDILI model is publicly available through https://github.com/TingLi2016/DeepDILI.

Developing novel computational prediction models for assessing chemical-induced neurotoxicity using naïve Bayes classifier technique

Development of reliable and efficient alternative in vivo methods for evaluation of the chemicals with potential neurotoxicity is an urgent need in the early stages of drug design. In this investigation, the computational prediction models for drug-induced neurotoxicity were developed by using the classical naïve Bayes classifier. Eight molecular properties closely relevant to neurotoxicity were selected. Then, 110 classification models were developed with using the eight important molecular descriptors and 10 types of fingerprints with 11 different maximum diameters. Among these 110 prediction models, the prediction model (NB-03) based on eight molecular descriptors combined with ECFP_10 fingerprints showed the best prediction performance, which gave 90.5% overall prediction accuracy for the training set and 82.1% concordance for the external test set. In addition, compared to naïve Bayes classifier, the recursive partitioning classifier displayed worse predictive performance for neurotoxicity. Therefore, the established NB-03 prediction model can be used as a reliable virtual screening tool to predict neurotoxicity in the early stages of drug design. Moreover, some structure alerts for characterizing neurotoxicity were identified in this research, which could give an important guidance for the chemists in structural modification and optimization to reduce the chemicals with potential neurotoxicity.

Developing novel in silico prediction models for assessing chemical reproductive toxicity using the naïve Bayes classifier method

Assessment of reproductive toxicity is one of the important safety considerations in drug development. Thus, in the present research, the naïve Bayes (NB)-classifier method was applied to develop binary classification models. Six important molecular descriptors for reproductive toxicity were selected by the genetic algorithm. Then, 110 classification models were developed using six molecular descriptors and10 types of fingerprints with 11 different maximum diameters. Among these established models, the model based on six molecular descriptors and the SciTegic extended-connectivity fingerprints with 20 maximum diameters (LCFC_20) displayed the best prediction performance for reproductive toxicity (NB-1), which gave a 0.884 receiver operating characteristic (ROC) score and 91.8% overall prediction accuracy for the Training Set, and produced a 0.888 ROC score and 83.0% overall accuracy for the external Test Set I. In addition, for the external rat multi-generation reproductive toxicity dataset (Test Set II), the NB-1 model generated a 0.806 ROC score and 85.1% concordance. The generated prediction results indicated that the NB-1 model could give robust and reliable predictions for a reproductive toxicity potential of chemicals. Thus, the established model could be applied to filter early-stage molecules for potential reproductive adverse effects. In addition, six important molecular descriptors and new structural alerts for reproductive toxicity were identified, which could help medicinal chemists rationally guide the optimization of lead compounds and select chemicals with the best prospects of being safe and effective.

In silico prediction of drug-induced developmental toxicity by using machine learning approaches

Some drugs and xenobiotics have the potential to disturb homeostasis, normal growth, differentiation, development or behavior during prenatal development or postnatally until puberty. Assessment of the developmental toxicity is one of the important safety considerations incorporated by international regulatory agencies. In this investigation, seven machine learning methods, including naïve Bayes, support vector machine, recursive partitioning, k-nearest neighbor, C4.5 decision tree, random forest and Adaboost, were used to build binary classification models for developmental toxicity. Among these models, the naïve Bayes classifier represented the best predictive performance and stability, which gave 91.11% overall prediction accuracy, 91.50% balanced accuracy and 0.818 MCC for the training set, and generated 83.93% concordance, 81.85% balanced accuracy and 0.627 MCC for the test set. The application domains were analyzed, and only one chemical in the test set was identified as outside the application domain. In addition, 10 important molecular descriptors related to developmental toxicity were selected by the genetic algorithm, which may contribute to explanation of the mechanisms of developmental toxicants. The best naïve Bayes classification model should be employed as alternative method for qualitative prediction of chemical-induced developmental toxicity in early stages of drug development.

Rapid Detection of Pesticide Residues in Paddy Water Using Surface-Enhanced Raman Spectroscopy

Pesticide residue in paddy water is one of the main factors affecting the quality and safety of rice, however, the negative effect of this residue can be effectively prevented and reduced through early detection. This study developed a rapid detection method for fonofos, phosmet, and sulfoxaflor in paddy water through chemometric methods and surface-enhanced Raman spectroscopy (SERS). Residue from paddy water samples was directly used for SERS measurement. The obtained spectra from the SERS can detect 0.5 mg/L fonofos, 0.25 mg/L phosmet, and 1 mg/L sulfoxaflor through the appearance of major characteristic peaks. Then, we used chemometric methods to develop models for the intelligent analysis of pesticides, alongside the SERS spectra. The classification models developed by K-nearest neighbor identified all of the samples, with an accuracy of 100%. For the quantitative analysis, the partial least squares regression models obtained the best predicted performance for fonofos and sulfoxaflor, and the support vector machine model provided optimal results, with a root-mean-square error of validation of 0.207 and a coefficient of determination of validation of 0.99952, for phosmet. Experiments for actual contaminated samples also showed that the above models predicted the pesticide residue values with high accuracy. Overall, using SERS with chemometric methods provided a simple and convenient approach for the detection of pesticide residues in paddy water.