Victor E. Kuz'min

Hierarchical QSAR technology based on the Simplex representation of molecular structure

This article is about the hierarchical quantitative structure–activity relationship technology (HiT QSAR) based on the Simplex representation of molecular structure (SiRMS) and its application for different QSAR/QSP(property)R tasks. The essence of this technology is a sequential solution (with the use of the information obtained on the previous steps) to the QSAR problem by the series of enhanced models of molecular structure description [from one dimensional (1D) to four dimensional (4D)]. It is a system of permanently improved solutions. In the SiRMS approach, every molecule is represented as a system of different simplexes (tetratomic fragments with fixed composition, structure, chirality and symmetry). The level of simplex descriptors detailing increases consecutively from the 1D to 4D representation of the molecular structure. The advantages of the approach reported here are the absence of “molecular alignment” problems, consideration of different physical–chemical properties of atoms (e.g. charge, lipophilicity, etc.), the high adequacy and good interpretability of obtained models and clear ways for molecular design. The efficiency of the HiT QSAR approach is demonstrated by comparing it with the most popular modern QSAR approaches on two representative examination sets. The examples of successful application of the HiT QSAR for various QSAR/QSPR investigations on the different levels (1D–4D) of the molecular structure description are also highlighted. The reliability of developed QSAR models as predictive virtual screening tools and their ability to serve as the base of directed drug design was validated by subsequent synthetic and biological experiments, among others. The HiT QSAR is realized as a complex of computer programs known as HiT QSAR software that also includes a powerful statistical block and a number of useful utilities.

Investigation of anticancer activity of macrocyclic Schiff bases by means of 4D-QSAR based on simplex representation of molecular structure

Influence of the molecular structure of macrocyclic pyridinophanes, their analogues and some other compounds on anticancer activity (Leukemia, central nervous system (CNS) cancer, prostate cancer, breast cancer, melanoma, non-small cell lung cancer, colon cancer, ovarian cancer, renal cancer) was investigated by means of a new 4D-QSAR approach based on the simplex representation of molecular structures (SiRMS). The number of group (N) is a tuning parameter which can be changed. As a rule For all the investigated molecules, the 3D structural models were first created and the set of conformers (fourth dimension) was used. Each conformer was represented as a system of different simplexes (tetratomic fragments of fixed structure, chirality and symmetry). Statistic characteristics of the QSAR partial least squares (PLS) models were satisfactory (correlation coefficient cross-validation coefficient ). The molecular fragments increasing and decreasing anticancer activity were defined. This information may be useful for the design and direct synthesis of novel anticancer agents.

Existing and Developing Approaches for QSAR Analysis of Mixtures

This review is devoted to the critical analysis of advantages and disadvantages of existing mixture descriptors and their usage in various QSAR/QSPR tasks. We describe good practices for the QSAR modeling of mixtures, data sources for mixtures, a discussion of various mixture descriptors and their application, recommendations about proper external validation specific for mixture QSAR modeling, and future perspectives of this field. The biggest problem in QSAR of mixtures is the lack of reliable data about the mixtures’ properties. Various mixture descriptors are used for the modeling of different endpoints. However, these descriptors have certain disadvantages, such as applicability only to 1 : 1 binary mixtures, and additive nature. The field of QSAR of mixtures is still under development, and existing efforts could be considered as a foundation for future approaches and studies. The usage of non‐additive mixture descriptors, which are sensitive to interaction effects, in combination with best practices of QSAR model development (e.g., thorough data collection and curation, rigorous external validation, etc.) will significantly improve the quality of QSAR studies of mixtures.

Interpretation of QSAR Models Based on Random Forest Methods.

A new algorithm for the interpretation of Random Forest models has been developed. It allows to calculate the contribution of each descriptor to the calculated property value. In case of the simplex representation of a molecular structure, contributions of individual atoms can be calculated, and thus it becomes possible to estimate the influence of separate molecular fragments on the investigated property. Such information can be used for the design of new compounds with a predefined property value. The proposed measure of descriptor contributions is not an alternative to the importance of Breiman’s variable, but it characterizes the contribution of a particular explanatory variable to the calculated response value.

QSPR Approach to Predict Nonadditive Properties of Mixtures. Application to Bubble Point Temperatures of Binary Mixtures of Liquids

This paper is devoted to the development of methodology for QSPR modeling of mixtures and its application to vapor/liquid equilibrium diagrams for bubble point temperatures of binary liquid mixtures. Two types of special mixture descriptors based on SiRMS and ISIDA approaches were developed. SiRMS‐based fragment descriptors involve atoms belonging to both components of the mixture, whereas the ISIDA fragments belong only to one of these components. The models were built on the data set containing the phase diagrams for 167 mixtures represented by different combinations of 67 pure liquids. Consensus models were developed using nonlinear Support Vector Machine (SVM), Associative Neural Networks (ASNN), and Random Forest (RF) approaches. For SVM and ASNN calculations, the ISIDA fragment descriptors were used, whereas Simplex descriptors were employed in RF models. The models have been validated using three different protocols: “Points out”, “Mixtures out” and “Compounds out”, based on the specific rules to form training/test sets in each fold of cross‐validation. A final validation of the models has been performed on an additional set of 94 mixtures represented by combinations of novel 34 compounds and modeling set chemicals with each other. The root mean squared error of predictions for new mixtures of already known liquids does not exceed 5.7 K, which outperforms COSMO‐RS models. Developed QSAR methodology can be applied to the modeling of any nonadditive property of binary mixtures (antiviral activities, drug formulation, etc.)

Universal Approach for Structural Interpretation of QSAR/QSPR Models

In this paper we offer a novel approach for the structural interpretation of QSAR models. The major advantage of our developed methodology is its universality, i.e., it can be applied to any QSAR/QSPR model irrespective of chemical descriptors and machine learning methods applied. This universality was achieved by using only the information obtained from substructures of the compounds of interest to interpret model outcomes. Reliability of the offered approach was confirmed by the results of three case studies, including end‐points of different types (continuous and binary classification) and nature (solubility, mutagenicity, and inhibition of Transglutaminase 2), various fragment and whole‐molecule descriptors (Simplex and Dragon), and multiple modeling techniques (partial least squares, random forest, and support vector machines). We compared the global contributions of molecular fragments obtained using our methodology with known SAR rules derived experimentally. In all cases high concordance between our interpretation and results published by others was observed. Although the proposed interpretation approach could be easily extended to any type of descriptors, we would recommend using Simplex descriptors to achieve a larger variety of investigated molecular fragments. The developed approach is a good tool for interpretation of such “black box” models like random forest, neural networks, etc. Analysis of fragment global contributions and their deviation across a dataset could be useful for the identification of key fragments and structural alerts. This information could be helpful to maximize the positive influence of structural surroundings on the given fragment and to decrease the negative effects.

Application of Quantum Chemical Approximations to Environmental Problems: Prediction of Water Solubility for Nitro Compounds

Water solubility values for 27 nitro compounds with experimentally measured values were computed using the conductor-like screening model for real solvent (COSMO-RS) based on the density functional theory and COSMO technique. We have found that the accuracy of the COSMO-RS approach for prediction of water solubility of liquid nitro compounds is impressively high (the errors are lower than 0.1 LU). However, for some solid nitro compounds, especially nitramines, there is sufficient disagreement between calculated and experimental values. In order to increase the accuracy of predictions the quantitative structure-property relationship (QSPR) part of the COSMO-RS approach has been modified. The solubility values calculated by the modified COSMO-RS method have shown much better agreement with the experimental values (the mean absolute errors are lower than 0.5 LU). Furthermore, this technique has been used for prediction of water solubility for an expanded set of 23 nitro compounds including nitroaromatic, nitramines, nitroanisoles, nitrogen rich compounds, and some their nitroso and amino derivatives with unknown experimental values. The solubility values predicted using the proposed computational technique could be useful for the determination of the environmental fate of military and industrial wastes and the development of remediation strategies for contaminated soils and waters. This predictive capability is especially important for unstable compounds and for compounds that have yet to be synthesized.

Application of Random Forest and Multiple Linear Regression Techniques to QSPR Prediction of an Aqueous Solubility for Military Compounds

The relationship between the aqueous solubility of more than two thousand eight hundred organic compounds and their structures was investigated using a QSPR approach based on Simplex Representation of Molecular Structure (SiRMS). The dataset consists of 2537 diverse organic compounds. Multiple Linear Regression (MLR) and Random Forest (RF) methods were used for statistical modeling at the 2D level of representation of molecular structure. Statistical characteristics of the best models are quite good (MLR method: R2=0.85, Q2=0.83; RF method: R2=0.99, R2oob=0.88). The external validation set of 301 compounds (including 47 nitro‐, nitroso‐ and nitrogen‐rich compounds of military interest) which were not included in the training set and modeling process, was used for evaluation of the models predictivity. Thus, well‐fitted and robust (R2test(MLR)=0.76 and R2test(RF)=0.82) models were obtained for both statistical techniques using descriptors based on the topological structural information only. The predicted solubility values for military compounds are in good agreement with experimental ones. Developed QSPR models represent powerful and easy‐to‐use virtual screening tool that can be recommended for prediction of aqueous solubility.

QSAR analysis of [(biphenyloxy)propyl]isoxazoles: agents against coxsackievirus B3

BACKGROUND Antiviral drugs are urgently needed for the treatment of acute and chronic diseases caused by enteroviruses such as coxsackievirus B3 (CVB3). The main goal of this study is quantitative structure-activity relationship (QSAR) analysis of anti-CVB3 activity (clinical CVB3 isolate 97927 [log IC50, µM]) and investigation of the selectivity of 25 ([biphenyloxy]propyl)isoxazoles, followed by computer-aided design and virtual screening of novel active compounds. DISCUSSION The 2D QSAR obtained models are quite satisfactory (R(2) = 0.84-0.99, Q(2) = 0.76-0.92, R(2)(ext) = 0.62-0.79). Compounds with high antiviral activity and selectivity have to contain 5-trifluoromethyl-[1,2,4]oxadiazole or 2,4-difluorophenyl fragments. Insertion of 2,5-dimethylbenzene, napthyl and especially biphenyl substituents into investigated compounds substantially decreases both their antiviral activity and selectivity. Several compounds were proposed as a result of design and virtual screening. A high level of activity of 2-methoxy-1-phenyl-1H-imidazo[4,5-c]pyridine (sm428) was confirmed experimentally. CONCLUSION Simplex representation of molecular structure allows successful QSAR analysis of anti-CVB3 activity of ([biphenyloxy]propyl)isoxazole derivatives. Two possible ways of battling CVB3 are considered as a future perspective.

New QSPR equations for prediction of aqueous solubility for military compounds

The development of a new quantitative structure-property relationship (QSPR) model to predict aqueous solubility (S(w)) accurately for compounds of military interest is presented. The ability of the new model to predict solubility is assessed and compared to available experimental data. A large set of structurally diverse organic compounds was used in this analysis. SiRMS methodology was employed to develop PLS models based on 135 training compounds and predictive accuracy was tested for 155 compounds selected for that purpose. The use of descriptors calculated only from the 2D level of representation of molecular structure produces a well-fitted and robust QSPR model (R(2)=0.90; Q(2)=0.87). Predictive ability for the model produced in this study on external test set (R(test)(2)=0.81) is comparable to the predictive ability of EPI Suite 4.0. Consensus solubility predictions using SiRMS and EPI models for 25 compounds of military interest (not included into the training set) have been completed.