Estimation of molecular properties by high-dimensional model representation

Additivity models have been widely employed to approximate unknown molecular properties based on previously measured or calculated data for similar molecules. This paper proposes an improved formulation of additivity, which is based on high-dimensional model representation (HDMR). HDMR is a general function-mapping technique that expresses the output of a multivariate system in terms of a hierarchy of cooperative effects among its input variables. HDMR rests on the general observation that, for many physical systems, only relatively low-order input variable cooperativity is significant. A molecule is expressed as a multivariate system by defining binary-valued input variables corresponding to the presence or absence of a chemical bond, with the molecular property as the output. Conventional additivity decomposes a molecular property into contributions from nonoverlapping subcomponents of fixed size. On the other hand, HDMR decomposes a molecular property into the exact contributions from the full hierarchy of its variable-sized subcomponents and contains additivity as a special case. The complete hierarchical structure of HDMR can in many cases lead to a much more accurate estimate than conventional additivity. Also, when full group additivity is not possible, HDMR gives an expression for a lower-order approximation for the missing group additivity value, greatly expanding the scope of HDMR compared to additivity. The component terms in an HDMR approximation have well-defined physical significance. Moreover, HDMR gives an exact expression for the truncation error in any given HDMR approximation, also with a well-defined physical significance. The HDMR model is tested for the enthalpy of formation of a broad range of organic molecules, and its advantages over additivity are illustrated.     

Quantitative Structure–Toxicity Relationships Using Tops-Mode. 1. Nitrobenzene Toxicity to Tetrahymena Pyriformis

Abstract

The TOPological Sub-Structural MOlecular DEsign (TOPS-MODE) approach (Estrada, E. SAR QSAR Environ. Res. 2000, 11, 55–73) has been introduced to the study of toxicological properties. The toxicity of 42 nitrobenzenes was studied with this approach obtaining a good quantitative structure–toxicity model. For the first time we compare the use of eight different weights in the diagonal entries of the bond matrix for selecting the best TOPS-MODE model. TOPS-MODE was used to derive the contribution of different fragments to the toxicity of studied compounds. These contributions were applied to calculate toxicity substituent constants for the groups present in the nitrobenzenes studied.     

神经网络方法在血管紧张素转换酶抑制剂定量构效关系建模中的应用

对20个ACEI化合物用量子化学方法进行结构优化并计算出10个参数,用9种不同隐含层节点数的BP神经网络研究了ACEI的定量构效关系,建立了节点为10/6/1的三层BP神经网络模型。结果表明:以量化理论计算所得参数可以构建合理的ACEI定量构效关系模型,神经网络模型M6的r2=0.995,S=0.050,6个验证集化合物的残差平方和为0.002,预测能力明显强于多元线形回归模型,亦优于同类文献报道,可作为ACEI研发领域中预测先导化合物活性的理论工具。     

Structure-activity relationships for non-congeneric structures. Application of substructural balance method for antiallergic activity

This paper presents a new research method of structure-activity relationships (SAR) based on the concept of substructural balance. By using antiallergic activity (PCA, rat, iv) of a non-congeneric set of 267 structures, the structural feature of active group is expressed in terms of substructural balance. Each structure was expressed with 100 new substructures and the number of each substructure in a molecule was counted. The substructural balance was expressed as their ratio. Structures were classified into three groups based on their potencies (ED50), active (44), median (33) and inactive (190) group. Using two substructural ratios, 80.53% of inactive and 57.58% of median structures were excluded from those that were active. Common features of active structures were shown as a zone indicating the optimal ranges of two substructural ratios. Two substructural ratios were determined out of 4950 substructural ratios, all possible combinations of 100 substructures (100C2), by selecting the greatest discriminatory power of inactive from active structures. The substructures used in this work include: the number of bonds comprising of the longest conjugate system, the number of skeletal atoms and the numbers of electron-donor pairs at certain distances in the molecule.     

Topological structural alerts modulations of mammalian cell mutagenicity for halogenated derivatives

Genotoxicity is a key toxicity endpoint for current regulatory requirements regarding new and existing chemicals. However, genotoxicity testing is time-consuming and costly, and involves the use of laboratory animals. This has motivated the development of computational approaches, designed to predict genotoxicity without the need to conduct laboratory tests. Currently, many existing computational methods, like quantitative structure–activity relationship (QSAR) models, provide limited information about the possible mechanisms involved in mutagenicity or predictions based on structural alerts (SAs) do not take statistical models into account. This paper describes an attempt to address this problem by using the TOPological Substructural MOlecular Design (TOPS-MODE) approach to develop and validate improved QSAR models for predicting the mutagenicity of a range of halogenated derivatives. Our most predictive model has an accuracy of 94.12%, exhibits excellent cross-validation and external set statistics. A reasonable interpretation of the model in term of SAs was achieved by means of bond contributions to activity. The results obtained led to the following conclusions: primary halogenated derivatives are more mutagenic than secondary ones; and substitution of chlorine by bromine increases mutagenicity while polyhalogenation decreases activity. The paper demonstrates the potential of the TOPS-MODE approach in developing QSAR models for identifying structural alerts for mutagenicity, combining high predictivity with relevant mechanistic interpretation.     

Fullerene quinazolinone conjugates targeting Mycobacterium tuberculosis: a combined molecular docking,QSAR, and ONIOM approach

Fullerene and its derivatives may bind to biological molecules, causing inhibitory effects. In this context, investigations of interactions of fullerene-based conjugates with proteins are of general interest. Particularly, fullerene and its derivatives demonstrate antibacterial properties; and one of the potential targets for drug design and health therapy is the inhibition of 6-oxopurine phosphoribosyltransferase in Mycobacterium tuberculosis (PDB code: 4RHY). In this article, the binding interactions between a series of quinazoline-4(3H)-ones and their fullerene derivatives with the target transferase were computationally investigated. Initially, we developed predictive quantitative structure-activity relationships (QSAR) models. Next, we introduced a simplified calculation schema that allows to evaluate relative binding affinities and to reveal specific mechanisms of action. For this purpose, the molecular docking approach was utilized to identify the native poses of the 18 transferase inhibitors. The binding pocket of the target protein was isolated and semi-empirical, and hybrid ONIOM scoring functions at different levels of theory were used to treat the ligands and the isolated binding pocket. The agreement within the calculated binding-free energies trends, as well as the agreement with the experimental data, suggests that the developed calculation schema can be used to estimate relative binding affinities towards 4RH. The combination of quantum-chemical models and QSAR models could be applied for future design of new selective inhibitors.     

Structure-Activity Relationships for Non-Congeneric Structures. Application of Substructural Balance Method for Antiallergic Activity

Abstract

This paper presents a new research method of structure-activity relationships (SAR) based on the concept of substructural balance. By using antiallergic activity (PCA, rat, iv) of a non-congeneri set of 267 structures, the structural feature of active group is expressed in terms of substructural balanance. Each structure was expressed with 100 new substructures and the number of each substructure in a molecule was counted. The substructural balance was expressed as their ratio. Structures were classified into three groups based on their potencies (ED₅₀), active (44), median (33) and inactactive (190) group. Using two substructural ratios, 80.53% of inactive and 57.58% of median structures were excluded from those that were active. Common features of active structures were shown as a zone indicating the optimal ranges of two substructural ratios. Two substructural ratios were determined out of 4950 substructural ratios, all possible combinations of 100 substructures (₁₀₀C₂), by selecting the greatest discriminatory power of inactive from active structures. The substructures used in this work include: the number of bonds comprising of the longest conjugate system, the number of skeletal atoms and the numbers of electron-donor pairs at certain distances in the molecule.     

Range,strength and anisotropy of intermolecular forces in atom–molecule systems: an atom–bond pairwise additivity approach

A new method is proposed to calculate bond energies and equilibrium distances in atom–molecule van der Waals complexes which arises from a balancing between long-range attraction and asymptotic tail of the repulsion. The method, based on correlation formulas between the polarizability of the interacting partners and the main interaction parameters, is an extension of an approach originally developed for atom–atom cases. The basic idea exploits the concept of bond polarizability additivity to represent both the molecular repulsion, in terms of a size which is mainly ascribed to the molecular bonds nearest to the probe atom, and the molecular attraction as due to multi-dispersion centers delocalized on the molecular frame. The method, mainly tested on hydrocarbon–rare gas complexes, can be considered as the starting point for the study of systems of higher complexity.     

Molecular structure–biological activity relationships on the basis of quantum-chemical calculations

Detailed quantum-chemical calculations by means of semiempirical all-valence electrons methods and a generalized (multivariable) rank correlation analysis are the fundamentals of a novel strategy of search for QSAR within homologous series of compounds. The set of molecular parameters (describing the electronic and conformational properties as well as potential interactions of the drugs) is calculated theoretically. Owing to the rank correlation method, no linear model (like LFER ) for the dependence of the biological activity upon the molecular parameters is presumed. The computed correlation coefficients are valued by carefully determined levels of statistical significance. Significant correlations are used to predict unknown activities in terms of ranks relative to the basic sample.     

TOPS-MODE based QSARs derived from heterogeneous series of compounds. Applications to the design of new herbicides

A new application of TOPological Sub-structural MOlecular DEsign (TOPS-MODE) was carried out in herbicides using computer-aided molecular design. Two series of compounds, one containing herbicide and the other containing nonherbicide compounds, were processed by a k-Means Cluster Analysis in order to design the training and prediction sets. A linear classification function to discriminate the herbicides from the nonherbicide compounds was developed. The model correctly and clearly classified 88% of active and 94% of inactive compounds in the training set. More specifically, the model showed a good global classification of 91%, i.e., (168 cases out of 185). While in the prediction set, they showed an overall predictability of 91% and 92% for active and inactive compounds, being the global percentage of good classification of 92%. To assess the range of model applicability, a virtual screening of structurally heterogeneous series of herbicidal compounds was carried out. Two hundred eighty-four out of 332 were correctly classified (86%). Furthermore this paper describes a fragment analysis in order to determine the contribution of several fragments toward herbicidal property; also the present of halogens in the selected fragments were analyzed. It seems that the present TOPS-MODE based QSAR is the first alternate general "in silico" technique to experimentation in herbicides discovery.     

Additive components of heteroatom influence in substituted alkanes. Polarization and depolarization of bonds

The general expression for the common one-electron density matrix (DM) of saturated organic molecules obtained previously in the framework of the Hückel type model (V. Gineityte, J. Mol. Struct. (Theochem) 343 (1995) 183) has been applied to reveal the additive components of the heteroatom influence in substituted alkanes. To this end, the occupation number of a basis orbital has been expressed as a sum of three terms describing the polarization and depolarization of bonds and the intramolecular charge transfer. These terms, in turn, have been related to certain types of direct (through-space) and indirect (through-bond) interactions of bond orbitals (BOs). In particular, changes in the secondary polarization of C-C and C-H bonds under the influence of a heteroatom giving rise to their induced dipole moments has been related to differences in the indirect interaction between the two BOs of the given bond before and after substitution. Additive quantum-chemical analogues of the classical inductive and electron-donating effects have been established. The above-mentioned expressions for the occupation numbers have been also applied to substantiate the implicit postulates of the classical chemistry about additivity of the heteroatom influence in substituted alkanes.