Global classification models for predicting acute toxicity of chemicals towards Daphnia magna.

Molecular descriptors reflecting structural information on hydrophobicity, reactivity, polarizability, hydrogen bond and charged groups, were used to predict the toxicity (pLC50) of chemicals towards Daphnia magna with global quantitative structure-activity/toxicity relationship (QSAR/QSTR) models. A sufficiently large dataset including 1517 chemical toxicity to Daphnia magna was divided into a training set (758 pLC50) and a test set (759 pLC50). By applying random forest algorithm, two classification models, Class Model A and Class Model B were developed, having prediction accuracy, sensitivity and specificity above 85% for Class 1 (with pLC50 ≤ 4.48) and Class 2 (with pLC50 > 4.48). The Class Model A was based on nine molecular descriptors and RF parameters of nodesize = 1, ntree = 80 and mtry = 2, and yielded accuracy of 92.3% (training set), 85.6% (test set) and 88.9% (total data set). Class Model B was based on ten descriptors and parameters, nodesize = 1, ntree = 90 and mtry = 2, produced accuracy of 88.3% (training set), 86.8% (test set) and 87.5% (total data set). The two classification models were satisfactory compared with other classification model reported in the literature, although classification models in this work dealt with more samples. Thus, the two classification models with a larger applicability domain provided efficient tools for assessing chemical aquatic toxicity towards Daphnia magna.

Large Dataset-Based Regression Model of Chemical Toxicity to Vibrio fischeri.

For the first time, a global regression quantitative structure-toxicity/activity relationship (QSTR/QSAR) model was developed for the toxicity of a large data set including 1236 chemicals towards Vibrio fischeri, by using random forest (RF) regression algorithm. The optimal RF model with RF parameters of mtry = 3, ntree = 150 and nodesize = 5 was based on 13 molecular descriptors. It can achieve accurate prediction for the toxicity of 99.1% of 1236 chemicals, and yield coefficients of determination R2 of 0.893 for 930 log(Mw/IBC50) in the training set, 0.723 for 306 log(Mw/IBC50) in the test se, and 0.865 for 1236 toxicity log(Mw/IBC50) in the total set. The optimal RF global model proposed in this work is comparable to other published local QSTR models on small datasets of the toxicity to Vibrio fischeri.

Support Vector Machine-Based Global Classification Model of the Toxicity of Organic Compounds to Vibrio fischeri.

Vibrio fischeri is widely used as the model species in toxicity and risk assessment. For the first time, a global classification model was proposed in this paper for a two-class problem (Class - 1 with log1/IBC50 ≤ 4.2 and Class + 1 with log1/IBC50 > 4.2, the unit of IBC50: mol/L) by utilizing a large data set of 601 toxicity log1/IBC50 of organic compounds to Vibrio fischeri. Dragon software was used to calculate 4885 molecular descriptors for each compound. Stepwise multiple linear regression (MLR) analysis was used to select the descriptor subset for the models. The ten molecular descriptors used in the classification model reflect the structural information on the Michael-type addition of nucleophiles, molecular branching, molecular size, polarizability, hydrophobic, and so on. Furthermore, these descriptors were interpreted from the point of view of toxicity mechanisms. The optimal support vector machine (SVM) model (C = 253.8 and γ = 0.009) was obtained with the genetic algorithm. The SVM classification model produced a prediction accuracy of 89.1% for the training set (451 log1/IBC50), of 80.0% for the test set (150 log1/IBC50), and of 86.9% for the total data set (601 log1/IBC50), which are higher than that (80.5%, 76%, and 79.4%, respectively) from the binary logistic regression (BLR) model. The global SVM classification model is successful, although it deals with a large data set in relation to the toxicity of organics to Vibrio fischeri.

Random forest algorithm-based accurate prediction of chemical toxicity to Tetrahymena pyriformis.

The random forest (RF) algorithm, together with ten Dragon descriptors, was used to develop a quantitative structure-toxicity/activity relationship (QSTR/QSAR) model for a larger data set of 1792 chemical toxicity pIGC50 towards Tetrahymena pyriformis. The optimal RF (ntree =300 and mtry =3) model yielded root mean square (rms) errors of 0.261 for the training set (1434 chemicals) and 0.348 for the test set (358 chemicals). Compared with other QSTR models reported in the literature, the optimal RF model in this paper is more accurate. The feasibility of applying the RF algorithm to predict chemical toxicity pIGC50 towards Tetrahymena pyriformis has been verified.

Random forest algorithm-based classification model of pesticide aquatic toxicity to fishes.

Aquatic toxicity of pesticides can result in poisoning of many non-target organisms, of which various fishes are the most prominent one. It is a challenge to predict the toxicity (LC50) classes of organic pesticides to various fish species from global QSAR models with a larger applicability domain. In this paper, by applying the random forest (RF) algorithm for a two-class problem, only eight molecular descriptors were used to develop a quantitative structure-activity relationship (QSAR) model for 1106 toxicity data (96 h, LC50) of organic pesticides to various fish species including Oncorhynchus mykiss, Lepomis macrochirus, Pimephales promelas, Brachydanio rerio, Cyprinodon, Cyprinus carpio, etc. By the prediction of the optimal RF Model I (ntree =280, mtry = 3 and nodesize = 5), the training set (885 organic pesticides) has the prediction accuracies of 99.6% for Class 1 (LC50 ≤ 10) and 96.7% for Class 2 (LC50 > 10); the test set (221 organic pesticides) has the accuracies being 90.8% for Class 1 and 91.2% for Class 2. The optimal RF Model I is satisfactory compared with other QSAR model reported in the literature, although its descriptor subset is small.

The Impact of Family Care for the Elderly on Women's Employment from the Perspective of Bargaining Power.

Due to the wishes of the elderly and the traditional family culture in China, family care is the main way of providing for the aged, and women's care is the main way. This is not conducive to the protection of women's employment rights and the realization of self-worth under the background of increasing women's autonomy. Based on the latest data of the China Health and Nutrition Survey Database (CHNS), this paper uses ordinary least squares (OLS) and the instrumental variable method of control endogeneity to analyze the influence of family care activities on the labor participation rate of married women. The innovation of this paper is to introduce family bargaining power into this kind of model for the first time, and further analyze the heterogeneity from the perspective of bargaining power differences. The empirical results show that the family elderly care activities have an obstacle effect on married women's participation in employment, and the family members with strong bargaining power will significantly hinder employment, so this paper puts forward policy recommendations in line with the actual situation, hoping to provide theoretical support for the improvement of the social security system for the elderly.

Correlation between the structure and skin permeability of compounds.

A three-descriptor quantitative structure-activity/toxicity relationship (QSAR/QSTR) model was developed for the skin permeability of a sufficiently large data set consisting of 274 compounds, by applying support vector machine (SVM) together with genetic algorithm. The optimal SVM model possesses the coefficient of determination R2 of 0.946 and root mean square (rms) error of 0.253 for the training set of 139 compounds; and a R2 of 0.872 and rms of 0.302 for the test set of 135 compounds. Compared with other models reported in the literature, our SVM model shows better statistical performance in a model that deals with more samples in the test set. Therefore, applying a SVM algorithm to develop a nonlinear QSAR model for skin permeability was achieved.

Support vector machine-based model for toxicity of organic compounds against fish.

Predicting the toxicity of chemicals to various fish species through chemometric approach is crucial for ecotoxicological assessment of existing as well as not yet synthesized chemicals. This paper reports a quantitative structure-activity/toxicity relationship (QSAR/QSTR) model for the toxicity pLC50 of organic chemicals against various fish species. Only six descriptors were used to develop the QSTR model, by applying support vector machine (SVM) together with genetic algorithm. The QSTR model was trained and established on a sufficiently large data set of 840 organic compounds and evaluated with a test set (281 compounds). Compared with other QSTRs reported in the literature, the optimal SVM model for fish toxicity produces better statistical results with determination coefficients R2 above 0.70 for both the training set and test set, although the QSTR model in this work possesses fewer molecular descriptors. Applying SVM and genetic algorithm to develop the QSTR model for pLC50 of organic compounds against various fish species is successful.

Prediction of reaction rate constants of hydroxyl radical with chemicals in water.

The rate constants (kOH ) of the reactions between organic micropollutants with hydroxyl radical (•OH) in aqueous systems are an important parameter to evaluate the persistence of organic compounds in the environment. In this paper, a support vector machine (SVM) model based on five descriptors was built to predict the reaction rate constants (log K = (log kOH )/MW ). The quantitative structure-activity relationship (QSAR) model of log K was obtained from a training set (600 compounds) and validated with a test set (395 compounds). The coefficients of determination R2 of the training and test sets are 0.923 and 0.925, respectively. The results suggest that the SVM model developed in this work possesses satisfactory prediction ability. PRACTITIONER POINTS: The rate constants of the reactions of organic micropollutants with •OH in aqueous systems were investigated. SVM model was established for the reaction rate constants (log K = (log kOH )/MW ). Only five molecular descriptors were used to predict 995 log K. A large data set was used for the test set (395 compounds).

Quantitative structure-toxicity relationships of organic chemicals against Pseudokirchneriella subcapitata.

Predicting the toxicity of organic toxicants to aquatic life through chemometric approach is challenging area. In this paper, a six-descriptor quantitative structure-activity/toxicity relationship (QSAR/QSTR) model was successfully developed for the toxicity pEC10 of organic chemicals against Pseudokirchneriella subcapitata, by applying support vector machine (SVM) together with genetic algorithm. A sufficiently large data set consisting of 334 organic chemicals was randomly divided into a training set (167 compounds) and a test set (167 compounds) with a ratio of 1:1. The optimal SVM model possesses coefficient of determination R2 of 0.76 and mean absolute error (MAE) of 0.60 for the training set and R2 of 0.75 and MAE of 0.61 for the test set. Compared with other models reported in the literature, our SVM model for the toxicity pEC10 shows significant statistical quality and satisfactory predictive ability, although it has fewer molecular descriptors and more samples in the test set. A QSTR model for pEC50 of organic chemicals against Pseudokirchneriella subcapitata was also developed with the same subsets and molecular descriptors.

Prediction of chemical toxicity to Tetrahymena pyriformis with four-descriptor models.

A quantitative structure-toxicity relationship (QSTR) model based on four descriptors was successfully developed for 1163 chemical toxicants against Tetrahymena pyriformis by applying general regression neural network (GRNN). The training set consisting of 600 organic compounds was used to train GRNN models that were evaluated with the test set of 563 compounds. For the optimal GRNN model, the training set possesses the coefficient of determination R2 of 0.86 and root mean square (rms) error of 0.41, and the test set has R2 of 0.80 and rms of 0.41. Investigated results indicate that the optimal GRNN model is accurate, although the GRNN model has only four descriptor and more samples in the test set.

Machine learning-based prediction of toxicity of organic compounds towards fathead minnow.

Predicting the acute toxicity of a large dataset of diverse chemicals against fathead minnows (Pimephales promelas) is challenging. In this paper, 963 organic compounds with acute toxicity towards fathead minnows were split into a training set (482 compounds) and a test set (481 compounds) with an approximate ratio of 1 : 1. Only six molecular descriptors were used to establish the quantitative structure-activity/toxicity relationship (QSAR/QSTR) model for 96 hour pLC50 through a support vector machine (SVM) along with genetic algorithm. The optimal SVM model (R 2 = 0.756) was verified using both internal (leave-one-out cross-validation) and external validations. The validation results (q int 2 = 0.699 and q ext 2 = 0.744) were satisfactory in predicting acute toxicity in fathead minnows compared with other models reported in the literature, although our SVM model has only six molecular descriptors and a large data set for the test set consisting of 481 compounds.

Prediction of Depuration Rate Constants for Polychlorinated Biphenyl Congeners.

The integral equation formalism polarizable continuum model (IEF-PCM) for solvent effects with the default solvent (water) and solvent parameters, together with the density functional theory method at 6-31G(d) level, was used to optimize molecular structures for polychlorinated biphenyl (PCB) congeners. Four molecular descriptors were selected to develop quantitative structure-activity relationship (QSAR) models for the depuration rate constants (k d) of 63 PCB congeners in a juvenile rainbow trout (Oncorhynchus mykiss). The optimal multiple linear regression (MLR) model has the correlation coefficient R of 0.933 and the root mean square (rms) error of 0.0681 for the total set of 63 PCB congeners. The support vector regression model has R of 0.953 and rms error of 0.0576 for the total set. Both the MLR and SVM QSAR models in this paper were accurate and acceptable compared with other QSAR models for the depuration rate of PCB congeners reported in references. Thus, applying IEF-PCM and B3LYP/6-31G(d) calculations for molecular descriptor derivation of PCB congeners is successful.

Coprecipitation-Gel Synthesis and Degradation Mechanism of Octahedral Li1.2Mn0.54Ni0.13Co0.13O2 as High-Performance Cathode Materials for Lithium-Ion Batteries.

The octahedral core-shell Li-rich layered cathode material of Li1.2Mn0.54Ni0.13Co0.13O2 can be synthesized via an ingenious coprecipitation-gel method without subsequent annealing. On the basis of detailed X-ray diffraction, scanning electron microscopy, transmission electron microscopy, and electron energy loss spectroscopy characterizations, it is suggested that the as-prepared material consists of an octahedral morphology and a new type of core-shell structure with a spinel-layered heterostructure inside, which is the result of overgrowth of the spinel structure with {111} facets on {001} facets of the layered structure in a single orientation. The surface area of Li1.2Mn0.54Ni0.13Co0.13O2 crystals where the spinel phase is located possesses sufficient Li and O vacancies, resulting in the reinsertion of Li into position after the first charge and maintenance of the interface stability via the replenishment of oxygen from the bulk region. Compared to that synthesized by the traditional coprecipitation method, the Li1.2Mn0.54Ni0.13Co0.13O2 synthesized by the coprecipitation-gel method exhibits higher discharge capacity and Coulombic efficiency, from 73.9% and 251.5 mAh g-1 for the spherical polycrystal material to 86.2% and 291.4 mAh g-1.

Novel Method for Structure-Activity Relationship of Aptamer Sequences for Human Prostate Cancer.

Prostate cancer (PCa) is one of the most common malignancies in men and seriously threatens men's health. Developing aptamer probes for PCa cells is of great significance for early diagnosis and treatment of PCa. This paper reports a classification model for SELEX-based aptamers, which were obtained with PCa cell line PCa-3M-1E8 (highly metastatic tumor cell) as target cells and PCa cell line PCa-3M-2B4 (low metastatic tumor cell) as control cells. On the basis of the SELEX principle, 100 oligonucleotide sequences from the 3rd round of SELEX were defined as low affinity and specificity aptamers, and 100 sequences from the 11th round were set as high affinity and specificity aptamers. Seven molecular descriptors were used for the classification model, which were calculated from amino acid sequences translated from DNA aptamer sequences with DNAMAN software. The classification model based on binary logical regression analysis has prediction accuracies, sensitivity, and specificity of about 80% for both the training set and test set. Therefore, it is feasible to calculate molecular descriptors from amino acid sequence translated from DNA aptamer sequences and develop a classification model for PCa cell line PCa-3M-1E8.

Pattern recognition of enrichment levels of SELEX-based candidate aptamers for human C-reactive protein.

Selecting aptamers for human C-reactive protein (CRP) would be of critical importance in predicting the risk for cardiovascular disease. The enrichment level of DNA aptamers is an important parameter for selecting candidate aptamers for further affinity and specificity determination. This paper is the first report on pattern recognition used for CRP aptamer enrichment levels in the systematic evolution of ligands by exponential enrichment (SELEX) process, by applying structure-activity relationship models. After generating 10 rounds of graphene oxide (GO)-SELEX and 1670 molecular descriptors, eight molecular descriptors were selected and five latent variables were then obtained with principal component analysis (PCA), to develop a support vector classification (SVC) model. The SVC model (C=8.1728 and γ=0.2333) optimized by the particle swarm optimization (PSO) algorithm possesses an accuracy of 88.15% for the training set. Prediction results of enrichment levels for the sequences with the frequencies of 6 and 5 are reasonable and acceptable, with accuracies of 70.59% and 76.37%, respectively.

Support vector machine classification of streptavidin-binding aptamers.

BACKGROUND: Synthesizing and characterizing aptamers with high affinity and specificity have been extensively carried out for analytical and biomedical applications. Few publications can be found that describe structure-activity relationships (SARs) of candidate aptamer sequences. METHODOLOGY: This paper reports pattern recognition with support vector machine (SVM) classification techniques for the identification of streptavidin-binding aptamers as "low" or "high" affinity aptamers. The SVM parameters C and γ were optimized using genetic algorithms. Four descriptors, the topological descriptor PW4 (path/walk 4--Randic shape index), the connectivity index X3A (average connectivity index chi-3), the topological charge index JGI2 (mean topological charge index of order 2), and the free energy E of the secondary structure, were used to describe the structures of candidate aptamer sequences from SELEX selection (Schütze et al. (2011) PLoS ONE (12):e29604). CONCLUSIONS: The predicted fractions of winning streptavidin-binding aptamers for ten rounds of SELEX conform to the aptamer evolutionary principles of SELEX-based screening. The feasibility of applying pattern recognition based on SVM and genetic algorithms for streptavidin-binding aptamers has been demonstrated.

A mathematical analysis of the selective enrichment of NECEEM-based non-SELEX.

Non-Systematic Evolution of Ligands by EXponential enrichment (SELEX)selection of aptamers, a novel technology for aptamer selection from libraries of random DNA (or RNA) sequences, involves repetitive steps of partitioning without polymerase chain reaction (PCR) amplification between them. This selection is based on non-equilibrium capillary electrophoresis of equilibrium mixtures (NECEEM) and has exceptionally high efficiency. In this paper, a mathematical analysis was carried out to predict the levels of enrichment of non-SELEX selection under different conditions such as different protein concentrations and different efficiencies of partitioning. Investigated results suggest that the magnitude of the bulk affinity (k d) being 10(4) or 10(5) µM for the initial pool has no obvious effect on selective enrichment and that the first, second, and third rounds of non-SELEX selection have different optimum protein concentration values [T f] that give maximum enrichment levels when [T f] ranges from 0.0005 to 0.5 µM. The significance of analyzing selective enrichment of NECEEM-based non-SELEX with the efficiency of partitioning target-bound ligands from free ligands has been demonstrated.

DFT-based theoretical QSPR models of Q-e parameters for the prediction of reactivity in free-radical copolymerizations.

Density functional theory (DFT) calculations at the B3LYP/6-31G(d) level were carried out for 47 vinyl monomers with structures C(1)H2 = C(2)HR3, and the calculated quantum chemical descriptors were used to construct quantitative structure-property relationship (QSPR) models of the reactivity parameters of monomers Q and e. Stepwise multiple linear regression analysis (MLRA) and artificial neural networks (ANN) were adopted to generate the models. Simulated with the final optimum back-propagation (BP) neural networks, the results show that predicted lnQ and e values are in good agreement with experimental data, with test sets possessing correlation coefficients of 0.982 for lnQ and 0.943 for e. The proposed ANN models have better prediction ability than existing models.

Prediction of refractive index of vinyl polymers by using density functional theory.

Density functional theory (DFT) calculations were carried out in the prediction of refractive index (n) of vinyl polymers at the B3LYP/6-31G(d) level. A set of quantum chemical descriptors calculated from monomers of polymers, the energy of the lowest unoccupied molecular orbital E(LUMO), the molecular average polarizability alpha, the heat capacity at constant volume C(v), and the most positive net atomic charge on hydrogen atoms in a molecule q(+), were used to built a general quantitative structure-property relationship (QSPR) model for refractive index. The proposed model gives the mean error of prediction 1.048% for the validation set, and has better prediction ability than the existing model.