支持向量机方法预测有机物的亨利常数

以有机物摩尔体积V、偶极项π*、氢键给予体的酸性am、氢键接受体的碱性βm等四种理化参数为输入变量,利用支持向量机方法对72种有机物的亨利常数值进行了定量预测研究。研究发现,采用支持向量机方法可以实现使用较少样本数据建模,并达到较好的预测结果。支持向量方法的预测结果远优于线性回归法预测结果。     

A support vector machine based method to predict success for polymerase chain reactions

Polymerase chain reaction (PCR) is one of the most popular molecular biological techniques and has been widely applied in many areas. However, PCR still faces challenges nowadays. During recent decades, the experimental procedure of PCR, including the primer design, was always the focus of attention, while little attention was paid to the analysis of the PCR template, and still nobody can accurately predict whether or not a DNA sequence can be simply amplified using conventional Taq DNA polymerase-based PCR protocol. In this study, we focus on the DNA template, the subject of PCR experiment, and introduce a support vector machine (SVM) based method to help evaluate PCR result. Through the Jackknife cross-validation test, our method achieves an accuracy of 92.06%, with 93.62% sensitivity and 90.53% specificity.     

Using support vector machine to predict beta- and gamma-turns in proteins

By using the composite vector with increment of diversity, position conservation scoring function, and predictive secondary structures to express the information of sequence, a support vector machine (SVM) algorithm for predicting beta- and gamma-turns in the proteins is proposed. The 426 and 320 nonhomologous protein chains described by Guruprasad and Rajkumar (Guruprasad and Rajkumar J. Biosci 2000, 25,143) are used for training and testing the predictive model of the beta- and gamma-turns, respectively. The overall prediction accuracy and the Matthews correlation coefficient in 7-fold cross-validation are 79.8% and 0.47, respectively, for the beta-turns. The overall prediction accuracy in 5-fold cross-validation is 61.0% for the gamma-turns. These results are significantly higher than the other algorithms in the prediction of beta- and gamma-turns using the same datasets. In addition, the 547 and 823 nonhomologous protein chains described by Fuchs and Alix (Fuchs and Alix Proteins: Struct Funct Bioinform 2005, 59, 828) are used for training and testing the predictive model of the beta- and gamma-turns, and better results are obtained. This algorithm may be helpful to improve the performance of protein turns' prediction. To ensure the ability of the SVM method to correctly classify beta-turn and non-beta-turn (gamma-turn and non-gamma-turn), the receiver operating characteristic threshold independent measure curves are provided.     

Support vector machines for the estimation of aqueous solubility

Support Vector Machines (SVMs) are used to estimate aqueous solubility of organic compounds. A SVM equipped with a Tanimoto similarity kernel estimates solubility with accuracy comparable to results from other reported methods where the same data sets have been studied. Complete cross-validation on a diverse data set resulted in a root-mean-squared error = 0.62 and R(2) = 0.88. The data input to the machine is in the form of molecular fingerprints. No physical parameters are explicitly involved in calculations.     

支持向量机法用于4-芳基-4氢-苯并比喃化合物的凋亡诱导活性研究

通过计算获得了39个4-芳基-4氢.苯并吡喃化合物的11个结构特征参数,应用逐步线性回归方法对参数进行筛选.并用支持向量回归(SVR)算法,对4-芳基-4氢-苯并吡喃化合物与其凋亡诱导活性进行定量构效关系(QSAR)研究.通过核函数参数的优化,建立了预测模型,训练集和预测集的实验值与计算值的平方相关系数R2分别为0.997和0.893.研究结果表明,支持向量回归算法可用于小样本数的药物分子设计研究,以合成具有更高生物活性的新药物.     

A comparison between neural network method and semi empirical equations to predict the solubility of different compounds in supercritical carbon dioxide

Accuracy of seven semi empirical equations for the estimation of solubility of 30 different compounds in supercritical carbon dioxide has been compared with a new neural network method. To base this comparison on a fair basis, a unique set of experimental data was used for both optimization of semi empirical equations’ parameters and training, validation and testing of neural network. Results showed that neural network method with an average relative deviation of about 5.3% was more accurate than the best semi empirical equation with an average relative deviation of about 15.96% for same compounds. It was also found that the average relative deviation of semi empirical equations varies sharply among different compounds, while this quantity is less dependent on material type for neural network method.     

Modifying gradient theory to predict the surface properties of halogenated hydrocarbons

A new method is proposed for predicting the surface tension, density profile, and thickness of the surface layer of a liquid near an interface using gradient theory. The objects of study are halogenated hydrocarbons. The algorithm for calculating surface properties includes a new modification of the Peng-Robinson cubic equation of state (EoS) that does not require information on the critical parameters, and a new procedure for calculating the influence parameter. Validation of the procedure for predicting the surface properties of liquids shows that the agreement between the calculated surface tension of halogenated hydrocarbons and the existing literature data is sufficient for practical use.     

多环芳烃水中溶解度的理论计算

建立了计算多环芳烃水中溶解度的数学表达式，用量子化学方法计算了7个多环芳烃的水中溶解度，计算结果与实验测定结果相符合．多环芳烃处于水体内体系状态能量愈高，其溶解度愈小，多环芳烃中的碳氢基团越多，溶解度越小．此时体系中的溶质呈单分子态，而不是聚集态．     

Methane conversion to C2 hydrocarbons in solid electrolyte membrane reactors

Solid electrolyte membrane reactors (SEMRs) have been used to both study and influence catalytic reaction rates. Methane coupling is the reaction most thoroughly and intensively studied in these membrane reactors. In the last 20 years, oxygen ion (O^2?), proton (H⁺) and mixed (O^2?-e^?, H⁺-e^?) conducting membranes have been tested in order to maximize the conversion of methane to C₂ compounds. The present review contains the fundamental operating principles of the various SEMR types and their applications in this reaction. The difficulties that should be overcome in order to promote this SEMR process to an industrial scale are discussed.     

Effect of bromine substitution on the solubility of gases in hydrocarbons and fluorocarbons

The interactions of oxygen and of carbon dioxide with normal and brominated octane are studied by analysing original experimental gas solubility data as a function of temperature in the range between 288 and 313 K. The temperature dependence of the solubility yields the thermodynamic properties of solvation. The influence of bromine substitution was studied by comparing the present data with that for perfluorooctane and 1-bromoperfluorooctane. A molecular interpretation of the effects observed was done using atomistic simulation. In order to simulate 1-bromooctane, parameters of the molecular force field were developped for bromoalkanes within the OPLS-AA framework. In general, simulation provides correct predictions of the solubility and of its temperature dependence, except in cases where values are too close (within the error bars of the simulation). Structural aspects of the solvation of the two gases were analysed in light of the site–site solute–solvant radial distribution functions. The relative importance of electrostatic interactions is assessed by modifying the intermolecular potential models for the gases.     

Models to predict solubility in ternary solvents based on sub-binary experimental data

The capability of the extended forms, of two well established cosolvency models, i.e. the combined nearly ideal binary solvent/Redlich-Kister equation and the modified Wilson model, used to predict the solute solubility in non-aqueous ternary solvent mixtures is presented. These predictions are based on the measured solubilities of anthracene in binary solvent mixtures. As a result the values of average percent deviations were less than 2% for the anthracene solubility in ternary mixtures. This work was also extended to other cosolvency models, ie. the extended Hildebrand solubility approach and the mixture response surface method, which are also commonly used for correlating solubility data in ternary solvents. The accuracy of the models is compared with each other and also with a published solubility model for ternary mixtures. The results illustrate that all models produced comparable accuracy.