基于分子参数的CYP2C9抑制剂的预测模型

细胞色素P4502C9(cytochrome P4502C9,CYP2C9)是肝脏重要的一种异物质代谢酶,许多药物或化学物质均可抑制和干扰其活性,在某种药物发现早期,预测基于CYP2C9抑制的药-药相互作用对筛选及发现新药具有重要意义。本文旨在建立CYP2C9抑制剂的预测模型,并确定抑制剂和非抑制剂显著不同的参数。选择81个化合物作为数据集,随机选其中64个为训练集,其余为验证集;选取250个分子参数给化合物数字化。采用逐步判别分析法(stepwise discriminant analysis method)和K-均值聚类分析法(K-Means cluster analysis method)模拟,建立数学模型,并用验证集检验模型的预测能力。结果表明:训练集的抑制剂正确率为96.4%,非抑制剂为97.2%;验证集的抑制剂正确率为85.7%,非抑制剂为90.0%。而采用K-均值聚类法时,抑制剂和非抑制剂的正确率也分别达到了82.9%和86.9%。对结果的深入分析找出对该模型贡献较大的参数为分子中氨基、烯基基团电拓扑状态指数、碳环数量以及疏水性参数,那些参数对区分抑制剂和非抑制剂两种结构差异、帮助指导CYP2C9抑制剂的筛选和发现具有重要意义。     

芳香/杂环磺胺类碳酸酐酶Ⅱ抑制剂的活性定量预测(英文)

对125个磺胺类碳酸酐酶Ⅱ抑制剂的生物活性进行了预测研究。利用ADRIANA.Code软件计算得到了化合物的一系列2D和3D结构描述符,从中选用了12个描述符进行建模。分别用数学随机划分的方法和Kohonen自组织神经网络的方法把数据集划分成两组不同的训练集和测试集。对于这两组不同的训练集和测试集,分别利用多元线性回归(MLR)和支持向量机(SVM)的方法进行建模,共得到4个模型。其中SVM得到的2个模型,训练集的相关系数在0.92以上,测试集预测的相关系数都在0.90以上。所有模型可进一步用于碳酸酐酶Ⅱ抑制剂的虚拟筛选。     

PhDD: A new pharmacophore-based de novo design method of drug-like molecules combined with assessment of synthetic accessibility

This account describes a new pharmacophore-based de novo design method of drug-like molecules (PhDD). The method PhDD first generates a set of new molecules that completely conform to the requirements of a given pharmacophore model, followed by a series of assessments to the generated molecules, including assessments of drug-likeness, bioactivity, and synthetic accessibility. PhDD is tested on three typical examples, namely, pharmacophore hypotheses of histone deacetylase (HDAC), cyclin-dependent kinase 2 (CDK2) and HIV-1 integrase (IN) inhibitors. The test results demonstrate that PhDD is able to generate molecules with novel structures but having similar biological functions with existing inhibitors. The validity of PhDD together with its ability of assessing synthetic accessibility makes it a useful tool in rational drug design.     

2D binary QSAR modeling of LPA3 receptor antagonism

A structurally diverse dataset of 119 compounds was used to develop and validate a 2D binary QSAR model for the LPA₃ receptor. The binary QSAR model was generated using an activity threshold of greater than 15% inhibition at 10 μM. The overall accuracy of the model on the training set was 82%. It had accuracies of 55% for active and 91% for inactive compounds, respectively. The model was validated using an external test set of 10 compounds. The accuracy on the external test set was 60% overall, identifying three out of seven actives and all three inactive compounds. This model was combined with similarity searching to rapidly screen libraries and select 14 candidate LPA₃ antagonists. Experimental assays confirmed 13 of these (93%) met the 15% inhibition threshold defining actives. The successful application of the model to select candidates for screening demonstrates the power of this binary QSAR model to prioritize compound selection for experimental consideration.     

环境化合物对鱼类毒性的定量构效关系研究

本研究基于定量构效关系方法预测环境化合物对鱼类的毒害(50%Lethal Concentration,LC50),并确定影响毒性关键分子的结构特征及几种模拟方法的比较.构建114个化学分子的数据集,随机选取85个75%分子为训练集,剩下的29个分子作为检验集,每个化学分子计算194个分子参数,分别采用逐步多元线性回归分析法(multiple linear regression,MLR)、主成分回归法(Principal Component Regression,PCA)和偏最小二乘法(Partial Least Square,PLS)构建定量结构-毒害关系(Quantitativestructure-activity relationships,QSTB)模型.用逐步多元线性回归分析法得出的训练集和预测集的实验值-logLC50与预测值-logLC50的相关系数分别为R2tr=0.86,R2te=O.83,说明该模型可靠性和鲁棒性较高;主成分回归法用8个主成分,其训练集和预测集的实验-logLC50与预测-logLC50的R2tr=0.81,R2te=O.77;偏最小二乘法用了5个潜变量,其训练集和预测集的实验-logLC50与预测-logLC50的R2tr=0.88,R2te=0.85.MLR方法得出化合物对鱼类的毒害影响较大的分子参数,主要分属电拓扑状态参数(SssO,SsCl,SdCH2,SsNH2)、分子连接指数(Xvo)以及修正Kappa指数(Ka2).以上研究对预测环境化合物的鱼类毒害(LC50),以及从机理上加深对有机物的毒性作用机理提供重要价值.     

Pharmacophore identification and virtual screening for methionyl-tRNA synthetase inhibitors

Aminoacyl-tRNA synthetases (aaRSs) are essential enzymes involved in protein biosynthesis in all living organisms and are an unexploited antibacterial targets, as many strains of bacteria have become resistant to all established classes of antibiotics. Therefore, the main aim of this study is to discover new lead molecules which would be useful as anti-bacterial compounds. Pharmacophore models were developed by using CATALYST HypoGen with a training set of 29 diverse methionyl-tRNA synthetase (MetRS) inhibitors. The best quantitative pharmacophore hypothesis (Hypo1) obtained a correlation coefficient of 0.975, root mean square deviation (RMSD) of 0.55 and cost difference (null cost-total cost) of 70.32. This Hypo1 was validated by two methods, first by using 104 test set molecules which resulted a correlation of 0.926 between HypoGen estimated activities versus experimental activities and secondly by Cat-Scramble validation method. This validated pharmacophore model was further used for screening databases for discovery of new MetRS inhibitors. The new lead compounds were further analyzed for drug-like properties. Homology modeled structure of Staphylococcus aureus MetRS was built and molecular docking studies were performed with many inhibitors using the newly built protein structure. Finally, it was found that the new leads exhibited good estimated inhibitory activity, calculated binding properties similar to experimentally proven compounds and also favorable drug-like properties.     

基于遗传算法的黄酮类化合物对PTKs抑制性的QSAR研究

用文献设定的结构参数和本文设定的结构参数,分别与由HyperChem7.5Student Evaluation计算得到的量化参数作为自变量构成2组数据,以逐步回归,遗传算法-偏最小二乘法(GA-PLS)和遗传算法-支持向量机(GA-SVM)等算法就黄酮类化合物对PTKs抑制性进行QSAR研究。用各算法模型处理数据,由本文设定的结构参数构成的数据集获得的预测结果更好,表明采用取代基团类型和取代位置结合的编码参数包含的信息更为丰富,对物质性质的描述更加合理。在各种算法中, GA-SVM模型均具有最佳预测效果,该算法对2组数据作留一法预测处理得到的相关系数R和PTKs抑制性实验值与预测值的平均绝对误差MAE分别为0.7595,0.2871和0.7864,0.2883。研究还表明,GA-PLS和GA-SVM联用算法的预测效果远高于单独使用的PLS和SVM算法;由逐步回归建立的MLR模型对2组数据进行计算处理,尽管拟合时相关系数R分别达到0.8136和0.8250,但作留一法交互验证时却下降到0.7113和0.7354,明显低于GA-PLS和GA-SVM联用算法。     

表皮生长因子受体酪氨酸激酶抑制剂的药效团研究

根据一系列表皮生长因子受体酪氨酸激酶抑制剂的三维定量构效关系研究,得到了该类抑制剂的药效团,研究结果与Novartis的药效团模型相当类似。药效团包括一个氢键受体,一个氢键给体,一个疏水区和一个带有氯或溴原子的苯环。该药效团对于研究表皮生长因子受体酪氨酸激酶抑制剂结构与活性的关系具有重要的意义。通过三维数据库搜索可能会得到新的先导化合物。     

Building a practical and reliable classifier for malware detection

Having a machine learning algorithm that can correctly classify malicious software has become a necessity as old methods of detection based on hashes and hand written heuristics tend to fail when dealing with the intensive flow of new malware. However, in order to be practical, the machine learning classifiers must also have a reasonable training time and a very small amount, preferably zero, of false positives. There were a few authors who addressed both these issues in their papers but creating such a model is more difficult when more than 3 million files are involved/needed in the training. We mapped a zero false positive perceptron in a new space, applied a feature selection algorithm and used the resulted model in an ensemble, voting or a rule based clustering system we’ve managed to achieve a detection rate around 99 % and 0.07 % false positives while keeping the training time suitable for large data sets.     

黄酮类化合物PIM-1激酶抑制活性的构效关系研究

为建立黄酮类化合物的PIM-1激酶抑制活性与其物化性质间的QSAR模型,本研究采用密度泛函理论(DFT)中的B3LYP方法,在6-311G**基组上全优化计算17个作为PIM-1激酶抑制剂的黄酮类化合物结构参数,运用SPSS 12.0 for Windows程序,将这些量子化参数作为理论描述符,逐步回归得到预测黄酮类化...