Prediction of n-octanol/water partition coefficients from PHYSPROP database using artificial neural networks and E-state indices

A new method, ALOGPS v 2.0 (http://www.lnh.unil.ch/~itetko/logp/), for the assessment of n-octanol/water partition coefficient, log P, was developed on the basis of neural network ensemble analysis of 12 908 organic compounds available from PHYSPROP database of Syracuse Research Corporation. The atom and bond-type E-state indices as well as the number of hydrogen and non-hydrogen atoms were used to represent the molecular structures. A preliminary selection of indices was performed by multiple linear regression analysis, and 75 input parameters were chosen. Some of the parameters combined several atom-type or bond-type indices with similar physicochemical properties. The neural network ensemble training was performed by efficient partition algorithm developed by the authors. The ensemble contained 50 neural networks, and each neural network had 10 neurons in one hidden layer. The prediction ability of the developed approach was estimated using both leave-one-out (LOO) technique and training/test protocol. In case of interseries predictions, i.e., when molecules in the test and in the training subsets were selected by chance from the same set of compounds, both approaches provided similar results. ALOGPS performance was significantly better than the results obtained by other tested methods. For a subset of 12 777 molecules the LOO results, namely correlation coefficient r(2)= 0.95, root mean squared error, RMSE = 0.39, and an absolute mean error, MAE = 0.29, were calculated. For two cross-series predictions, i.e., when molecules in the training and in the test sets belong to different series of compounds, all analyzed methods performed less efficiently. The decrease in the performance could be explained by a different diversity of molecules in the training and in the test sets. However, even for such difficult cases the ALOGPS method provided better prediction ability than the other tested methods. We have shown that the diversity of the training sets rather than the design of the methods is the main factor determining their prediction ability for new data. A comparative performance of the methods as well as a dependence on the number of non-hydrogen atoms in a molecule is also presented.     

Model selection based on structural similarity-method description and application to water solubility prediction

A method is introduced that allows one to select, for a given property and compound, among several prediction methods the presumably best-performing scheme based on prediction errors evaluated for structurally similar compounds. The latter are selected through analysis of atom-centered fragments (ACFs) in accord with a k nearest neighbor procedure in the two-dimensional structural space. The approach is illustrated with seven estimation methods for the water solubility of organic compounds and a reference set of 1876 compounds with validated experimental values. The discussion includes a comparison with the similarity-based error correction as an alternative approach to improve the performance of prediction methods and an extension that enables an ad hoc specification of the application domain.     

Target family-directed exploration of scaffolds with different SAR profiles

The scaffold concept is widely applied in chemoinformatics and medicinal chemistry to organize bioactive compounds according to common core structures or associate compound classes with specific biological activities. A variety of scaffold analyses have been carried out to derive statistics for scaffold distributions, generate structural organization schemes, or identify scaffolds that preferentially occur in given compound activity classes. Herein we further extend scaffold analysis by identifying scaffolds that display defined SAR profiles consisting of multiple properties. A structural relationship-based scaffold network has been designed as the basic data structure underlying our analysis. From network representations of scaffolds extracted from compounds active against 32 different target families, scaffolds with different SAR profiles have been extracted on the basis of decision trees that capture structural and functional characteristics of scaffolds in different ways. More than 600 scaffolds and 100 scaffold clusters were assigned to 10 SAR profiles. These scaffold sets represent different activity and target selectivity profiles and are provided for further SAR investigations including, for example, the exploration of alternative analog series for a given target of target family or the design of novel compounds on the basis of scaffold(s) with desired SAR profiles.     

SURFCOMP: a novel graph-based approach to molecular surface comparison

Analysis of the distributions of physicochemical properties mapped onto molecular surfaces can highlight important similarities or differences between compound classes, contributing to rational drug design efforts. Here we present an approach that uses maximal common subgraph comparison and harmonic shape image matching to detect locally similar regions between two molecular surfaces augmented with properties such as the electrostatic potential or lipophilicity. The complexity of the problem is reduced by a set of filters that implement various geometric and physicochemical heuristics. The approach was tested on dihydrofolate reductase and thermolysin inhibitors and was shown to recover the correct alignments of the compounds bound in the active sites.     

In silico identification of natural products with anticancer activity using a chemo-structural database of Brazilian biodiversity

Cancer is one of the leading causes of death worldwide, and the number of patients has only increased each year, despite the considerable efforts and investments in scientific research. Since natural products (NPs) may serve as suitable sources for drug development, the cytotoxicity against cancer cells of 2221 compounds from the Nuclei of Bioassays, Ecophysiology, and Biosynthesis of Natural Products Database (NuBBE_DB) was predicted using CDRUG algorithm. Molecular modeling, chemoinformatics, and chemometric tools were then used to analyze the structural and physicochemical properties of these compounds. We compared the positive NPs with FDA-approved anticancer drugs and predicted the molecular targets involved in the anticancer activity. In the present study, 46 families comprising potential anticancer compounds and at least 19 molecular targets involved in oncogenesis. To the best of our knowledge, this is the first large-scale study conducted to evaluate the potentiality of NPs sourced from Brazilian biodiversity as anticancer agents, using in silico approaches. Our results provided interesting insights about the mechanism of action of these compounds, and also suggested that their structural diversity may aid structure-based optimization strategies for developing novel drugs for cancer therapy.     

Rapid Identification of Potential Drug Candidates from Multi-Million Compounds’ Repositories. Combination of 2D Similarity Search with 3D Ligand/Structure Based Methods and In Vitro Screening

Rapid in silico selection of target focused libraries from commercial repositories is an attractive and cost-effective approach in early drug discovery. If structures of active compounds are available, rapid 2D similarity search can be performed on multimillion compounds’ databases. This approach can be combined with physico-chemical parameter and diversity filtering, bioisosteric replacements, and fragment-based approaches for performing a first round biological screening. Our objectives were to investigate the combination of 2D similarity search with various 3D ligand and structure-based methods for hit expansion and validation, in order to increase the hit rate and novelty. In the present account, six case studies are described and the efficiency of mixing is evaluated. While sequentially combined 2D/3D similarity approach increases the hit rate significantly, sequential combination of 2D similarity with pharmacophore model or 3D docking enriched the resulting focused library with novel chemotypes. Parallel integrated approaches allowed the comparison of the various 2D and 3D methods and revealed that 2D similarity-based and 3D ligand and structure-based techniques are often complementary, and their combinations represent a powerful synergy. Finally, the lessons we learnt including the advantages and pitfalls of the described approaches are discussed.     

运用模糊神经网络表达和预测链烷烃pVT性质

采用一种基于遗传算法的新型模糊神经网络方法研究链烷烃类化合物的pVT性质。该方法综合神经网络、遗传算法与模糊系统三种柔性智能计算技术的优点,具有良好的学习能力,不易陷入局部最小区域,学习速度较快,网络知识以模糊语言变量的形式加以表达,易于理解。用分子连接性指数对24种链烷烃化合物结构和pVT数据进行学习,进而预测另外14种未知化合物的pVT性质,较好地揭示出化合物分子结构与pVT性质之间的关系,并给出了良好的关联与预测结果。     

Structural diversity of small molecule libraries

A novel method for assessing structural diversity is presented. Maximum common subgraph identity is used as the measure of similarity between two chemical structures. A conditional probability treatment of similarity distributions for libraries of chemical structures is used to define diversity. This evaluation method together with the evaluation of traditional physicochemical properties is used to assess a large number of chemical libraries and to understand structural differences between these.