Lidija R. Jevrić

Non-linear assessment of anticancer activity of 17-picolyl and 17-picolinylidene androstane derivatives--chemometric guidelines for further syntheses.

The present paper deals with prediction of cytotoxic activity of 17-picolyl and 17-picolinylidene androstane derivatives toward androgen receptor negative prostate cancer cell line (PC-3). The prediction was achieved applying artificial neural networks (ANNs) method on the basis of molecular descriptors. The most important descriptors (skin permeability (SP), Madin-Darby canine kidney cell permeability (MDCK) and universal salt solubility factor (S+SF)) were selected by using stepwise selection coupled with partial least squares method. The ANN modelling was carried out in order to obtain reliable models which can facilitate further synthesis of androstane derivatives with high antiproliferative activity toward PC-3 cell line. The modelling procedure resulted in three ANN models with the best statistical performance. The obtained results show that the established ANN models can be applied for required purpose.

Chemometric estimation of the RP TLC retention behaviour of some estrane derivatives by using multivariate regression methods

AbstractQuantitative structure-retention relationship (QSRR) was developed for a series of estrane derivatives, on the basis of their retention data, obtained in reversed-phase thin-layer chromatography (RP TLC), and in silico molecular descriptors. Physicochemical and topological descriptors, as well as molecular bulkiness descriptors, were calculated from the optimized molecular structures. Full geometry optimization was achieved by using Austin Model 1 (AM1) semi-empirical molecular orbital method. In the present study, QSRR analysis was based on principal component analysis (PCA), multiple linear regression (MLR) and partial least squares (PLS) method. PCA was applied in order to reveal similarities or dissimilarities between analytes, and MLR and PLS regression methods were carried out in order to identify the most important in silico molecular descriptors and quantify their influence on the retention behaviour of studied compounds. Physically meaningful and statistically significant structure-retention relationships were established.

Assessment of Chromatographic Lipophilicity of Some Anhydro-D-Aldose Derivatives on Different Stationary Phases by QSRR Approach

This study is based on the analysis of the retention behavior of some anhydro-D-aldose derivatives in reversed-phase thin-layer chromatography (RP-TLC). Chromatographic separations were achieved applying a methanol/water mobile phase and three different stationary phases: C18-bonded silica gel, silica gel impregnated with paraffin oil, and silica gel impregnated with squalane. Retention behavior of the analyzed molecules was defined by constant and correlated with in silico molecular descriptors. According to the statistical validation parameters, obtained results showed that the established multiple-linear quantitative structure-retention relationship models (MLR-QSRR) can successfully predict the chromatographic lipophilicity of structurally similar compounds. In the present study the influence of some functional groups on the total lipophilicity of studied derivatives was determined as well. Principal component analysis (PCA) was applied in order to obtain an overview of similarity or dissimilarity among the analyzed compounds and hierarchical cluster analysis (HCA) was applied in order to compare separation characteristics of the applied stationary phases.

Chemometric guidelines for selection of cultivation conditions influencing the antioxidant potential of beetroot extracts

Display Omitted 64 beetroot samples were cultivated under different conditions.Novel approach to samples similarity analysis was applied.The differences between the samples were detected.ANN-R model best predicts antioxidant activity of beetroot extracts. The estimation of influences of different cultivation conditions on chemical composition of crops can be quite complex task. Usually, the classical statistic approach (comparison of statistical characteristics of different populations) is applied. However, this study presents the analysis of the influence of different cultivation conditions (cultivation with or without foil cover, different preceding crops and fertilization) on the contents of dry matter, betalains and phenolic compounds in 64 beetroot (Beta vulgaris L. ssp. vulgaris) samples and their antioxidant potential (IC50) by using powerful chemometric tools, such as Wald-Wolfowitz run test, cluster analysis and sum of ranking differences, in order to obtain groupings of the samples which share similar properties or to test if the samples come from the same population. Besides the aforementioned classification or pattern recognition analysis, the regression methods (linear, multiple linear and artificial neural network regressions) were carried out in order to establish reliable linear and non-linear relationships between IC50 and contents of betalains, phenolic compounds and dry matter. The results reveal certain influence of foil cover, used during the cultivation, on beetroot antioxidant potential and content of the determined betalains and phenolic compounds, as well as the strong non-linear relationship between the analyzed variables. Certain influence of preceding crops on phenolic compounds content is detected by Wald-Wolfowitz run test.

Reversed-phase thin-layer chromatography behavior of aldopentose derivatives

Quantitative structure-retention relationships (QSRR) have been used to study the chromatographic behavior of some aldopentose. The behavior of aldopentose derivatives was investigated by means of reversed-phase thin-layer chromatography (RP TLC) on silica gel impregnated with paraffin oil stationary phases. Binary mixtures of methanol-water, acetone- water and dioxane-water were used as mobile phases. Retention factors, RM 0, corresponding to zero percent organic modifier in the aqueous mobile phase was determined. Lipophilicity, C0, was calculated as the ratio of the intercept and slope values. There was satisfactory correlation between them and log P values calculated using different theoretical procedures. Some of these correlations offer very good predicting models, which are important for a better understanding of the relationships between chemical structure and retention. The study showed that the hydrophobic parameters RM 0 and C0 can be used as a measures of lipophilicity of investigated compounds.

How to rank and discriminate artificial neural networks? Case study: prediction of anticancer activity of 17-picolyl and 17-picolinylidene androstane derivatives

Model discrimination is still not a resolved task. The classical statistical approaches lead to different results (for the same models) and at the same time a lot of models seem to be statistically equivalent. The authors deliberately select such conditions when their algorithm is superior. Hence, it is better to apply different approaches to compare and rank the models fairly. This paper presents the application of methodology called sum of ranking differences (SRD) to rank the artificial neural network models [quantitative structure–activity relationship (QSAR) models] designed for prediction of anticancer activity of 17-picolyl and 17-picolinylidene androstane derivatives toward androgen receptor negative prostate cancer cells (AR-, PC-3). The SRD method suggests the consistent models, in terms of compounds order and proximity to the golden standard, which should preferably be used in the prediction of anticancer activity of studied androstane derivatives.

Preselection of A- and B- modified d-homo lactone and d-seco androstane derivatives as potent compounds with antiproliferative activity against breast and prostate cancer cells – QSAR approach and molecular docking analysis

The problem with trial-and-error approach in organic synthesis of targeted anticancer compounds can be successfully avoided by computational modeling of molecules, docking studies and chemometric tools. It has been proven that A- and B- modified d-homo lactone and d-seco androstane derivatives are compounds with significant antiproliferative activity against estrogen-independent breast adenocarcinoma (ER-, MDA-MB-231) and androgen-independent prostate cancer cells (AR-, PC-3). This paper presents the quantitative structure-activity relationship (QSAR) models based on artificial neural networks (ANNs) which are able to predict whether d-homo lactone and/or d-seco androstane-based compounds will express antiproliferative activity against breast cancer cells (MDA-MB-231) or not. Also, the present paper describes the molecular docking study of 3β-acetoxy-5α,6α-epoxy- (3) and 6α,7α-epoxy-1,4-dien-3-one (24) d-homo lactone androstane derivatives, as well as 4-en-3-one (15) d-seco androstane derivative, which are compounds with strong or moderate antiproliferative activity against prostate cancer cells (PC-3), and compares them with commercially available medicament for prostate cancer - abiraterone. The obtained promising results can be used as guidelines in further syntheses of novel d-homo lactone and d-seco androstane derivatives with antiproliferative activity against breast and prostate cancer cells.

Comprehensive QSRR modeling as a starting point in characterization and further development of anticancer drugs based on 17α-picolyl and 17(E)-picolinylidene androstane structures

The selection of the most promising anticancer compounds from the pool of the huge number of synthesized molecules is a quite complex task. There are many compounds characterization approaches which can suggest the best structural features of a molecule with the highest antiproliferative effect on the certain type of cancer cell lines. One of these approaches is the lipophilicity determination of compounds and the analysis of its correlation with the anticancer activity. Since the importance of the lipophilicity is underlined in many earlier studies, this study is focused on determination of lipophilicity of previously synthesized 17α-picolyl and 17(E)-picolinylidene androstane derivatives by using reversed-phase high performance liquid chromatography (RP-HPLC) as a very fast, effective and relatively cheap method. Determination of the chromatographic lipophilicity of the studied androstanes can be considered as the part of their physicochemical characterization, which is a very important step in their further selection as drug candidates. The present study does not neglect the in silico approach. The determined chromatographic lipophilicity was analyzed by quantitative structure-retention relationship (QSRR) approach in order to reveal which molecular characteristics contribute mostly to the typical behavior of the androstanes in the applied chromatographic system, and thus to their lipophilicity. Classical statistical approach and Sum of Ranking Differences method were used for selection of the best QSRR models which should be used in prediction of chromatographic lipophilicity of studied androstane derivatives.

Structure-Retention Analysis of Some 1,6-anhydrohexose and D-aldopentose Derivatives by Linear Multivariate Approach

Quantitative structure-retention relationships (QSRR) was developed for eleven 1,6-anhydrohexose and D-aldopentose derivatives using their retention data (RM0 values), obtained by normal phase (NP) thin-layer chromatography (TLC) and partition coefficients for n-octanol/water bi-phase system–logP. In order to select molecular descriptors that best describe the retention behavior of the investigated molecules, principal component analysis (PCA) was carried out, followed by hierarchical cluster analysis (HCA) and multiple linear regression (MLR). The best QSRR models were validated by leave-one-out technique and by the calculation of statistical parameters. Established mathematical models are meaningful and statistically significant and can successfully predict retention behavior of investigated 1,6-anhydrohexose and D-aldopentose derivatives.

Chemometric approach to texture profile analysis of kombucha fermented milk products

In the present work, relationships between the textural characteristics of fermented milk products obtained by kombucha inoculums with various teas were investigated by using chemometric analysis. The presented data which describe numerically the textural characteristics (firmness, consistency, cohesiveness and index of viscosity) were analysed. The quadratic correlation was determined between the textural characteristics of fermented milk products obtained at fermentation temperatures of 40 and 43 °C, using milk with 0.8, 1.6 and 2.8% milk fat and kombucha inoculums cultivated on the extracts of peppermint, stinging nettle, wild thyme and winter savory. Hierarchical cluster analysis (HCA) was performed to identify the similarities among the fermented products. The best mathematical models predicting the textural characteristics of investigated samples were developed. The results of this study indicate that textural characteristics of sample based on winter savory have a significant effect on textural characteristics of samples based on peppermint, stinging nettle and wild thyme, which can be very useful in the determination of products texture profile.