Using Molecular Quantum Similarity Measures as Descriptors in Quantitative Structure-Toxicity Relationships

Abstract

In this paper molecular quantum similarity measures (MQSM) are used to describe molecular toxicity and to construct Quantitative Structure-Toxicity Relationships (QSTR) models. This study continues the recently described relationships between MQSM and log P values, which permits to use the theoretical MQSM as an alternative to the empirical hydrophobic parameter in QSPR studies. In addition a new type of MQSM is presented in this work: it is based on the expectation value of electron–electron repulsion energy. The molecular properties studied here, as application examples are aquatic toxicity, toxicology on Bacteria and inhibition of a macromolecule employing four different molecular sets.     

Using molecular quantum similarity measures under stochastic transformation to describe physical properties of molecular systems

The application of molecular quantum similarity measures (MQSM) to correlate physicochemical properties is reported. Satisfactory quantitative structure-property relationship (QSPR) models are obtained for three molecular sets, where boiling points and chromatographic retention times and indices are studied. In this work, MQSM are scaled using a stochastic transformation and related to molecular properties using the partial least-squares technique.     

Molecular quantum similarity matrix based clustering of molecules using dendrograms

A new scheme for general classification of quantum objects is presented. Based on molecular quantum similarity matrices (MQSM), different algorithms are presented for generating Molecular Quantum Similarity Dendrograms (MQSD). An application of MQSD is presented for a set of steroid molecules.     

基于QSAR模型的有机磷化合物毒性预测

随着有机磷化合物(OPs)的广泛应用,其在越来越多的环境介质中被检测出来。大多数OPs具有毒性,但人们缺乏快速且有效的预测手段来对毒性进行评估。本文将结合E-Dragon软件计算的分子描述符,采用不同的QSAR模型对36个OPs的毒性进行预测。文中采用后退法作为描述符筛选方法,以均方根误差(RMSE)作为评价标准,共找到14个对线性核函数支持向量机(SVM)模型贡献较大的描述符;在最终得到的SVM模型交叉验证结果中,计算值与实际值的相关系数为0. 913,均方根误差为0. 388;外部测试验证结果中,平均相对误差为9. 10%。此外,采用多元线性回归(MLR)、人工神经网络(ANN)以及偏最小二乘回归(PLS)模型对OPs的毒性进行预测,交叉验证结果显示,三个模型的计算值与实际值的相关系数分别为0. 878、0. 686与0. 620,没有SVM模型的预测能力好。因此采用线性核函数的SVM模型对OPs进行毒性预测是一个行之有效的方法。     

Synthetic and structural carboxylate chemistry of neurotoxic aluminum in relevance to human diseases

The contact of Al(III) with biological components in human physiology has increased significantly over the years, due to a number of factors, prominent among which stands the rapid acidification of the environment and the concomitant introduction of that abundant metal ion in human biological fluids. As a result, pathophysiological aberrations in humans have arisen due to Al(III) (neuro)toxicity. Among the efforts targeting the elucidation of the factors responsible for Al(III) toxicity is the exploration of the requisite Al(III)-carboxylate chemistry in aqueous media, and its relevance to soluble, potentially bioavailable species capable of exerting toxic effects. A detailed synthetic, structural, and spectroscopic account of the Al(III)-carboxylate complexes, purported to exist as components in aqueous Al(III)-carboxylic acid speciation, is presented. The structures described are classified as mononuclear, dinuclear, trinuclear, tetranuclear, and polynuclear species, arising from various aqueous and non-aqueous Al(III)-carboxylate ligand reactions. Moreover, the solution chemistry and kinetic behavior of the so far reported complexes is given, with the specific aim of comparing their solid state and solution chemical and structural properties. In this sense, a comprehensive picture on the Al(III) speciation, in the presence of various physiological or biologically relevant carboxylate ligands, appears to emerge, which is expected to contribute to the understanding of Al(III) (neuro)toxicity and its consequence(s) in a multitude of human diseases. Carboxylate containing low and high molecular mass components stand prominent in their chemical preference to react with Al(III) in biological fluids. In this context, factors considered to influence the aqueous low molecular mass Al(III)-carboxylate chemistry, thus affecting the solubility and possibly the bioavailability of the resulting species, are discussed as potential research links to the ultimate manifestation of Al(III) toxicity at the cellular level.     

应用分子全息QSAR预测多溴二苯醚(PBDEs)的毒性

多溴二苯醚(PBDEs)可能会激活芳香烃受体的信号传导通路, 从而对人类和野生动物的健康产生负面影响. 鉴于多溴二苯醚实验毒性数据有限, 发展基于结构的化合物毒性预测模型具有重要的实际意义. 本文基于一种新的分子结构表征方法—— 分子全息, 研究了18种多溴二苯醚结构与毒性之间的关系, 建立了相关性显著、稳健性强的QSAR模型(r²= 0.991, q²_LOO= 0.917). 随机选出14种多溴二苯醚为训练集, 其他4种化合物为测试集以验证分子全息QSAR模型的稳健性和预测能力. 结果在最佳建模条件下得到模型的统计参数如下：r² = 0.988, q²_LOO = 0.598, r²_pred = 0.955, 预测值与实验值之间的均方根误差(RMSE)为0.155. 这表明基于分子全息的QSAR模型可以对多溴二苯醚毒性进行比较准确的预测. 本文同时利用分子全息QSAR模型色码图, 探讨了影响多溴二苯醚毒性的分子结构特征及分子机理.     

An Algorithm Framework for Drug-Induced Liver Injury Prediction Based on Genetic Algorithm and Ensemble Learning

In the process of drug discovery, drug-induced liver injury (DILI) is still an active research field and is one of the most common and important issues in toxicity evaluation research. It directly leads to the high wear attrition of the drug. At present, there are a variety of computer algorithms based on molecular representations to predict DILI. It is found that a single molecular representation method is insufficient to complete the task of toxicity prediction, and multiple molecular fingerprint fusion methods have been used as model input. In order to solve the problem of high dimensional and unbalanced DILI prediction data, this paper integrates existing datasets and designs a new algorithm framework, Rotation-Ensemble-GA (R-E-GA). The main idea is to find a feature subset with better predictive performance after rotating the fusion vector of high-dimensional molecular representation in the feature space. Then, an Adaboost-type ensemble learning method is integrated into R-E-GA to improve the prediction accuracy. The experimental results show that the performance of R-E-GA is better than other state-of-art algorithms including ensemble learning-based and graph neural network-based methods. Through five-fold cross-validation, the R-E-GA obtains an ACC of 0.77, an F1 score of 0.769, and an AUC of 0.842.     

Impact of graphene-based nanomaterials (GBNMs) on the structural and functional conformations of hepcidin peptide

Graphene-based nanomaterials (GBNMs) are widely used in various industrial and biomedical applications. GBNMs of different compositions, size and shapes are being introduced without thorough toxicity evaluation due to the unavailability of regulatory guidelines. Computational toxicity prediction methods are used by regulatory bodies to quickly assess health hazards caused by newer materials. Due to increasing demand of GBNMs in various size and functional groups in industrial and consumer based applications, rapid and reliable computational toxicity assessment methods are urgently needed. In the present work, we investigate the impact of graphene and graphene oxide nanomaterials on the structural conformations of small hepcidin peptide and compare the materials for their structural and conformational changes. Our molecular dynamics simulation studies revealed conformational changes in hepcidin due to its interaction with GBMNs, which results in a loss of its functional properties. Our results indicate that hepcidin peptide undergo severe structural deformations when superimposed on the graphene sheet in comparison to graphene oxide sheet. These observations suggest that graphene is more toxic than a graphene oxide nanosheet of similar area. Overall, this study indicates that computational methods based on structural deformation, using molecular dynamics (MD) simulations, can be used for the early evaluation of toxicity potential of novel nanomaterials.     

Antilymphoma Effect of Incomptine A: In Vivo,In Silico,and Toxicological Studies

Incomptine A (IA) is a sesquiterpene lactone isolated from Decachaeta incompta that induces apoptosis, reactive oxygen species production, and a differential protein expression on the U-937 (diffuse histiocytic lymphoma) cell line. In this work, the antitumor potential of IA was investigated on Balb/c mice inoculated with U-937 cells and through the brine shrimp lethality (BSL) test. Furthermore, IA was subjected to molecular docking study using as targets proteins associated with processes of cancer as apoptosis, oxidative stress, and glycolytic metabolism. In addition to determining the potential toxicity of IA in human, its acute toxicity was performed in mice. Results reveals that IA showed high antilymphoma activity and BSL with an EC₅₀ of 2.4 mg/kg and LC₅₀ 16.7 µg/mL, respectively. The molecular docking study revealed that IA has strong interaction on all targets used. In the acute oral toxicity, IA had a LD₅₀ of 149 mg/kg. The results showed that the activities of IA including antilymphoma activity, BSL, acute toxicity, and in silico interactions were close to the methotrexate, an anticancer drug used as positive control. These findings suggest that IA may serve as a candidate for the development of a new drug to combat lymphoma.     

Qsar Models for Predicting the Acute Toxicity of Selected Organic Chemicals with Diverse Structures to Aquatic Non-Vertebrates and Humans

Abstract

The linear and non-linear relationships of acute toxicity (as determined on five aquatic non-vertebrates and humans) to molecular structure have been investigated on 38 structurally-diverse chemicals. The compounds selected are the organic chemicals from the 50 priority chemicals prescribed by the Multicentre Evaluation of In Vitro Cytotoxicity (MEIC) programme. The models used for the evaluations are the best combination of physico-chemical properties that could be obtained so far for each organism, using the partial least squares projection to latent structures (PLS) regression method and backpropagated neural networks (BPN). Non-linear models, whether derived from PLS regression or backpropagated neural networks, appear to be better than linear models for describing the relationship between acute toxicity and molecular structure. BPN models, in turn, outperform non-linear models obtained from PLS regression. The predictive power of BPN models for the crustacean test species are better than the model for humans (based on human lethal concentration). The physico-chemical properties found to be important to predict both human acute toxicity and the toxicity to aquatic non-vertebrates are the n?octanol water partition coefficient (Pow) and heat of formation (HF). Aside from the two former properties, the contribution of parameters that reflect size and electronic properties of the molecule to the model is also high, but the type of physico-chemical properties differs from one model to another. In all of the best BPN models, some of the principal component analysis (PCA) scores of the ¹³C-NMR spectrum, with electron withdrawing/accepting capacity (LUMO, HOMO and IP) are molecular size/volume (VDW or MS1) parameters are relevant. The chemical deviating from the QSAR models include non-pesticides as well as some of the pesticides tested. The latter type of chemical fits in a number of the QSAR models. Outliers for one species may be different from those of other test organisms.     

Holographic quantitative structure-activity relationship for prediction of the toxicity of polybrominated diphenyl ether congeners

Polybrominated diphenyl ether congeners (PBDEs) might activate the AhR (aromatic hydrocarbon receptor) signal transduction, and thus might have an adverse effect on the health of humans and wildlife. Because of the limited experimental data, it is important and necessary to develop structure-based models for prediction of the toxicity of the compounds. In this study, a new molecular structure representation, molecular hologram, was employed to investigate the quantitative relationship between toxicity and molecular structures for 18 PBDEs. The model with the significant correlation and robustness (r ² = 0.991, q ² _LOO = 0.917) was developed. To verify the robustness and prediction capacity of the derived model, 14 PBDEs were randomly selected from the database as the training set, while the rest were used as the test set. The results generated under the same modeling conditions as the optimal model are as follows: r ² = 0.988, q ² _LOO = 0.598, r ² _pred = 0.955, and RMSE (root-mean-square of errors) = 0.155, suggesting the excellent ability of the derived model to predict the toxicity of PBDEs. Furthermore, the structural features and molecular mechanism related to the toxicity of PBDEs were explored using HQSAR color coding.