Search

The validation principles on Quantitative Structure Activity Relationship issued by Organization for Economic and Co‐operation and Development describe three criteria of model assessment: goodness of fit, robustness and prediction. In the case of robustness, two ways are possible as internal validation: bootstrap and cross‐validation. We compared these validation metrics by checking their sample size dependence, rank correlations to other metrics and uncertainty. We used modeling methods from multivariate linear regression to artificial neural network on 14 open access datasets. We found that the metrics provide similar sample size dependence and correlation to other validation parameters. The individual uncertainty originating from the calculation recipes of the metrics is much smaller for both ways than the part caused by the selection of the training set or the training/test split. We concluded that the metrics of the two techniques are interchangeable, but the interpretation of cross‐validation parameters is easier according to their similar range to goodness‐of‐fit and prediction metrics. Furthermore, the variance originating from the random elements of the calculation of cross‐validation metrics is slightly smaller than those of bootstrap ones, if equal calculation load is applied. [ABSTRACT FROM AUTHOR]

The article first utilizes the hyperspectral image grayscale, texture information, and reflectance spectral information at characteristic wavelengths to establish the corresponding quantitative analysis models of nutrient content in maize plants using three modeling methods, MLR, PCR, and PLS. Then from the basis of polarized reflection features, it was inferred that the polarization degree features at sensitive wavelengths were extracted using polarized spectra and combined with chemometric techniques to achieve a quantitative analysis of the degree of nutrient stress in maize. Finally, the feature variables extracted on the hyperspectral and polarized-reflection spectral measurement systems were fused with multiple information. A diagnostic evaluation model of fertilizer utilization with polarization-hyperspectral multidimensional light information was established. The results showed that for the new slow-release fertilizers, SF1 and YNPK had higher nitrogen utilization rates, 9.91% and 7.43% higher than N1PK, respectively. And the nitrogen fertilizer utilization rate was correspondingly higher by 6%-7% in 2020 than in 2019 for each fertilizer application treatment.

The quantitative structure-property relationship QSPR method using Multiple Linear Regression MLR and Partial Least Squares PLS methodologies was performed for 160 organic compounds (hydrocarbons, branched alkanes, branched and unbranched alkenes, and alkynes). The MLR and PLS methods were employed to explore the correlation between the molecular descriptors which are the structural representation while the critical temperature Tc is the property representation. Using Dragon descriptors, this study was aimed at developing a predictive and robust QSPR models for predicting Tc. According to the squared correlation coefficients (R² =0.942 and 0.941), standard error (s =0.88 and 0.797) and the leave-one-out cross-validation correlation coefficients (Q²Loo = 0.834 and 0.932), for the MLR and PLS methods respectively, the results demonstrated almost identical qualities and good predictive ability for both the MLR and PLS models. [ABSTRACT FROM AUTHOR]

As they facilitate the cleavage of single and double stranded DNA to relax supercoils, unwind catenanes, and condense chromosomes in eukaryotic cells, Topoisomerase plays crucial roles in cellular reproduction and DNA organization. Because the unrepaired single and double stranded DNA breaks these complexes generate might result in apoptosis and cell death, they are cytotoxic agents.In this study, 28 compounds derived from phosphorus-substituted quinoline are subjected to a quantitative structure–activity relationship (QSAR) using partial least squares, principal component regression, and multiple linear regression. The Gaussian 09 software and the Molecular Operating Environment program were used to calculate molecular descriptors. The anti-proliferative activity was correlated with a variety of electronic and structural characteristics of the molecules, such as EHOMO (energy of the highest occupied molecular orbital) and ELUMO (energy of the lowest unoccupied molecular orbital), which provided evidence for the modeling. The B3LYP/6-31G (d, p) level of theory's Density Functional Theory (DFT) approach was used to compute these electronic properties, and Principal Component Analysis (PCA) was used to test for collinearity between the descriptors. In fact, three alternative prediction models were created using various 2D and 3D descriptor counts, and they were each assessed using the statistical metrics of coefficient of determination (R2) and root mean squared error (RMSE). A MLR model had the best predictive performance of all the constructed models, as indicated by R2 and RMSE of 0.865 and 0.316, respectively.Three proteins (6G77, 2NS2, and 5K47) for lung, ovarian, and kidney malignancies showed strong binding affinities via hydrophobic interactions and H-bonds with the pertinent chemicals by crystal structure modeling. Compounds C11, C19 and C26, respectively, showed the highest binding energy for ovarian, kidney and lung cancer. The outcomes of the molecular dynamic MD simulation diagram were produced to support the molecular docking findings from earlier research, which demonstrated that inhibitors were stable in the active sites of the selected proteins for 10 ns. This raises the possibility that these chemicals could serve as a valuable model for the development and synthesis of more effective anticancer prospects.

Angiotensin-converting-enzyme inhibitors (ACEI) are a group of drugs primarily used in the treatment of cardiovascular disease. In this study, MLR and PLS QSAR models were developed to evaluate the antihypertensive activity of new enalapril analogs, while binding affinities between each analog and ACE were determined with AutoDock. Consequently, analogs presenting para and meta trifluoromethyl substitutions, and a N,N-dialkyl aliphatic amide were the most promising analogs, exhibiting an IC50 of 0.009 nM and affinity energies of −8.9 kcal/mol, surpassing those of enalapril. Furthermore, all promising analogs were predicted to be less toxic than enalapril according to the software PreADMET.

Objectives: Present communication deals with two- and three-dimensional (2D and 3D) QSAR studies of twenty saponin analogue for antifungal activity against Candida albicans. Methods: The 2D-QSAR model for the prediction was obtained by applying Multiple Linear Regression (MLR) method, giving r² = 0.8551 and q² = 0.7717 and Partial Least Squares (PLS) method, giving r² = 0.8551 and q² = 0.7717. 3D-QSAR study was performed using the stepwise variable selection k-nearest neighbour molecular field analysis (kNN-MFA) approach for both electrostatic and steric fields. Two different kNN-MFA methods (SA and GA) were used for the building of 3D-QSAR models. Results: The best model shows interesting result in terms of internal (q² > 0.62) and external (predictive r² > 0.52) predictivity for training and test set. Conclusion: Thus, QSAR models showed that hydrophobic and electrostatic effects dominantly determine the binding affinities. Hence the QSAR models proposed in this work would be further useful for development of new antifungal agents from medicinal plants and can help in the design of novel potent molecule. [ABSTRACT FROM AUTHOR]

The quantitative structure–activity relationship (QSAR) analyses were carried out in a series of novel sulfonamide derivatives as the procollagen C-proteinase inhibitors for treatment of fibrotic conditions. Sphere exclusion method was used to classify data set into categories of train and test set at different radii ranging from 0.9 to 0.5. Multiple linear regression (MLR), principal component regression (PCR) and partial least squares (PLS) were used as the regression methods and stepwise, Genetic algorithm (GA), and simulated annealing (SA) were used as the feature selection methods. Three of the statistically best significant models were chosen from the results for discussion. Model 1 was obtained by MLR–SA methodology at a radius of 1.6. This model with a coefficient of determination (r2) = 0.71 can well predict the real inhibitor activities. Cross-validated q2 of this model, 0.64, indicates good internal predictive power of the model. External validation of the model (pred_r2 = 0.85) showed that the model can well predict activity of novel PCP inhibitors. The model 2 which developed using PLS–SW explains 72% (r2 = 0.72) of the total variance in the training set as well as it has internal (q2) and external (pred_r2) predictive ability of ∼67% and ∼71% respectively. The last developed model by PCR–SA has a correlation coefficient (r2) of 0.68 which can explains 68% of the variance in the observed activity values. In this case internal and external validations are 0.61 and 0.75, respectively. Alignment Independent (AI) and atomic valence connectivity index (chiv) have the greatest effect on the biological activities. Developed models can be useful in designing and synthesis of effective and optimized novel PCP inhibitors which can be used for treatment of fibrotic conditions.

Protein kinases are essential components of various signaling pathways and represent attractive targets for therapeutic interventions. Kinase inhibitors are currently used to treat malignant tumors, as well as autoimmune diseases, due to their involvement in immune cell signaling. In this study, three-dimensional quantitative structure–activity relationship (3D-QSAR) analyses, including Multiple Linear Regression (MLR), Partial Least Squares (PLS), Multiple Non-Linear Regression (MNLR), Artificial Neural Network (ANN) and cross-validation analyses, were performed on a set of P38 MAP kinases as anti-cancer agents. This method, which is based on molecular modeling (molecular mechanics, Hartree-Fock (HF)), was used to determine the structural parameters, electronic properties, and energy associated with the molecules we examined. MLR, PLS, and MNLR analyses were performed on 46 protein P38 MAP kinase analogs to determine the relationships between molecular descriptors and the anti-cancer properties of the P38 MAP kinase analogs. The MLR model was validated by the external validation and standardization approach. The ANN, given the descriptors obtained from the MLR, exhibited a correlation coefficient close to 0.94. The predicted model was confirmed by two methods, leave-one-out (LOO) cross-validation and scrambling (or Y-randomization). We observed a high correlation between predicted and experimental activity, thereby both validating and demonstrating the high quality of the QSAR model that we described.

Neonicotinoids are the fastest-growing class of insecticides successfully applied in plant protection, human and animal health care. The significant resistance increases led to the urgent need for alternative new neonicotinoids, with improved insecticidal activity. We performed molecular docking to describe a common binding mode of neonicotinoids into the nicotinic acetylcholine receptor, and to select the appropriate conformations to derive models. These were further used in a QSAR study employing both linear and nonlinear approaches to model the inhibitory activity against the Cowpea aphids. Linear modeling was performed by multiple linear regression and partial least squares and nonlinear modeling by artificial neural networks and support vector machine methods. The OECD principles were considered for QSAR models validation. Robust models with predictive power were found for neonicotinoid diverse structures. Based on our QSAR and docking outcomes, five new insecticides were predicted, according to the model applicability domain, the ligand efficiencies, and the binding mode. Therefore, the developed models can be confidently used for the prediction of the insecticidal activity of new chemicals, saving a substantial amount of time and money and, also, contributing to the chemical risk assessment. [ABSTRACT FROM AUTHOR]

To address the issue of unavailability of selective JNK1 inhibitors resulting in off-target effects, leading to multiple diseases, an endeavour to discover novel and specific JNK1 inhibitors is taken up in the present study. To achieve this goal, computer-aided drug design approach has been used and a validated 2D QSAR model, of excellent statistical quality, has been developed through MLR (Multiple Linear Regression) and PLS (Partial Lest Square) method. The r² value obtained through PLS method (0.97) corroborated the r² value that has been obtained through MLR approach (0.97). The insights obtained through in-depth study of the developed model has capacitated us to design optimized molecules (compound A1OOP and compound A2SSR) with better selectivity profile than the most active compound of the selected set of compounds that have been employed to build the QSAR model. Additionally, molecular docking and structure based pharmacophore design have been performed to ensure that the affinity of the designed molecules towards JNK1 receptor and evaluation of their ADME properties have been done to ensure their lead likeness. Further, extremely small values for Tanimoto similarity index are obtained that clearly suggest that the designed molecules are novel. [ABSTRACT FROM AUTHOR]

In this study, clarithromycin, amoxicillin, and lansoprazole agents for the treatment of Helicobacter pylori, which has a role in chronic active gastritis and peptic ulcer disease, has been determinated by the simultaneous spectrophotometric method. Multivariable calibration methods - chemometric methods have been successfully applied to the obtained data. Benefiting from the relationship between the absorbance and the concentration, with the aid of the synthetic mixture, in the urine sample was subjected to the determination of each active substance and the data were analyzed by MLR (Multi Linear Regression), PLS (Partial Least Squares Method) and PCR (Basic Component Regression). The MLR (Multiple Linear Regression), PLS (Partial Least Squares Method) and PCR (Principal Component Regression) are the methods of chemometric calibrations and the results obtained with these methods are extremely fast and most reliable. [ABSTRACT FROM AUTHOR]

In this study, we investigated the structure-activity relationships of a series of β-carboline alkaloid derivatives using the 2D-QSAR and molecular docking, in order to identify the mode of interaction between β-carboline derivatives and the PLK1 kinase, and determine their key substituents responsible for the cytotoxic activity. The obtained QSAR models using multiple linear regression (MLR) and partial least squares (PLS) methods showed a high correlation between the experimental activity and the predicted one by PLS (R2PLS = 0.82, q2 = 0.72) and MLR (R2MLR = 0.82, q2 = 0.72). An external dataset was used to test the extrapolation power of the models which resulted in an R2PLS (EV) = 0.76; RMSE = 0.39. The 2D-QSAR analysis reveals that lipophilicity plays an important role in the cytotoxic activity of this group of β-carboline derivatives. Indeed, the molecular docking study into the active site of the polo-like kinase (PLK1) revealed that the most active ligand 57 shows higher binding energy and interacts, especially by H-bonds and hydrophobic interactions, with the active site of the PLK1 kinase. Consequently, the results obtained from the 2D-QSAR and docking studies provided a useful tool to design new and potent β-carboline derivatives as cytotoxic agents. [ABSTRACT FROM AUTHOR]

Prediction performance often depends on the cross- and test validation protocols applied. Several combinations of different cross-validation variants and model-building techniques were used to reveal their complexity. Two case studies (acute toxicity data) were examined, applying five-fold cross-validation (with random, contiguous and Venetian blind forms) and leave-one-out cross-validation (CV). External test sets showed the effects and differences between the validation protocols. The models were generated with multiple linear regression (MLR), principal component regression (PCR), partial least squares (PLS) regression, artificial neural networks (ANN) and support vector machines (SVM). The comparisons were made by the sum of ranking differences (SRD) and factorial analysis of variance (ANOVA). The largest bias and variance could be assigned to the MLR method and contiguous block cross-validation. SRD can provide a unique and unambiguous ranking of methods and CV variants. Venetian blind cross-validation is a promising tool. The generated models were also compared based on their basic performance parameters (r2 and Q2). MLR produced the largest gap, while PCR gave the smallest. Although PCR is the best validated and balanced technique, SVM always outperformed the other methods, when experimental values were the benchmark. Variable selection was advantageous, and the modelling had a larger influence than CV variants. [ABSTRACT FROM AUTHOR]

Prediction performance often depends on the cross- and test validation protocols applied. Several combinations of different cross-validation variants and model-building techniques were used to reveal their complexity. Two case studies (acute toxicity data) were examined, applying five-fold cross-validation (with random, contiguous and Venetian blind forms) and leave-one-out cross-validation (CV). External test sets showed the effects and differences between the validation protocols. The models were generated with multiple linear regression (MLR), principal component regression (PCR), partial least squares (PLS) regression, artificial neural networks (ANN) and support vector machines (SVM). The comparisons were made by the sum of ranking differences (SRD) and factorial analysis of variance (ANOVA). The largest bias and variance could be assigned to the MLR method and contiguous block cross-validation. SRD can provide a unique and unambiguous ranking of methods and CV variants. Venetian blind cross-validation is a promising tool. The generated models were also compared based on their basic performance parameters (r2 and Q2). MLR produced the largest gap, while PCR gave the smallest. Although PCR is the best validated and balanced technique, SVM always outperformed the other methods, when experimental values were the benchmark. Variable selection was advantageous, and the modelling had a larger influence than CV variants. [ABSTRACT FROM AUTHOR]

The inhibition ability of six bis-benzimidazole derivatives against corrosion of mild steel in acidic medium was modeled by the electronic and lipophilic descriptors such as: the energy of the highest occupied molecular orbital (EHOMO), the energy of the lowest unoccupied molecular orbital (ELUMO), the dipole moment (μ), the molecular critical volume (Vc), the inhibitor concentration (Ci), the logarithm of the partition coefficient (log P), and the molecular mass (M), by means of quantitative structure property relationships (QSPR). To do so, the structure electronic properties of these molecules were investigated by using three quantum methods: a semi-empirical method (AM1), the Hartree-Fock method (HF) and the density functional theory method (DFT), at B3LYP/6-311G level of theory. The correlation was developed by three mathematical models, based-QSPR approaches: the multiple linear regression (MLR), the multiple polynomial regression (MPR) and the partial least squares regression (PLS). A very good determination coefficient (R² = 0.99), adjusted determination coefficient (R² adj = 0.99) and predicted determination coefficient (R2 pred = 0.97), were obtained. The calculated inhibition efficiency increased in the order OBBI > HBBI > BBI > MBBI> BBBI > EBBI, was in good agreement with experimental results. [ABSTRACT FROM AUTHOR]

As they facilitate the cleavage of single and double stranded DNA to relax supercoils, unwind catenanes, and condense chromosomes in eukaryotic cells, Topoisomerase plays crucial roles in cellular reproduction and DNA organization. Because the unrepaired single and double stranded DNA breaks these complexes generate might result in apoptosis and cell death, they are cytotoxic agents. In this study, 28 compounds derived from phosphorus-substituted quinoline are subjected to a quantitative structure–activity relationship (QSAR) using partial least squares, principal component regression, and multiple linear regression. The Gaussian 09 software and the Molecular Operating Environment program were used to calculate molecular descriptors. The anti-proliferative activity was correlated with a variety of electronic and structural characteristics of the molecules, such as E HOMO (energy of the highest occupied molecular orbital) and E LUMO (energy of the lowest unoccupied molecular orbital), which provided evidence for the modeling. The B3LYP/6-31G (d, p) level of theory's Density Functional Theory (DFT) approach was used to compute these electronic properties, and Principal Component Analysis (PCA) was used to test for collinearity between the descriptors. In fact, three alternative prediction models were created using various 2D and 3D descriptor counts, and they were each assessed using the statistical metrics of coefficient of determination (R2) and root mean squared error (RMSE). A MLR model had the best predictive performance of all the constructed models, as indicated by R2 and RMSE of 0.865 and 0.316, respectively. Three proteins (6G77, 2NS2, and 5K47) for lung, ovarian, and kidney malignancies showed strong binding affinities via hydrophobic interactions and H-bonds with the pertinent chemicals by crystal structure modeling. Compounds C11 , C19 and C26 , respectively, showed the highest binding energy for ovarian, kidney and lung cancer. The outcomes of the molecular dynamic MD simulation diagram were produced to support the molecular docking findings from earlier research, which demonstrated that inhibitors were stable in the active sites of the selected proteins for 10 ns. This raises the possibility that these chemicals could serve as a valuable model for the development and synthesis of more effective anticancer prospects. [ABSTRACT FROM AUTHOR]

Herein we have studied the cytotoxicity and quantitative structure–activity relationship (QSAR) of heterocyclic compounds containing cyclic urea and thiourea nuclei. A set of 22 hydantoin and thiohydantoin related heterocyclic compounds were investigated with respect to their LC50 values (Log of LC50) against brine shrimp lethality bioassay in order to derive the 2D-QSAR models using MLR, PLS and ANN methods. The best predictive models by MLR, PLS and ANN methods gave highly significant square correlation coefficient (R2) values of 0.83, 0.81 and 0.91 respectively. The model also exhibited good predictive power confirmed by the high value of cross validated correlation coefficient Q2 (0.74).

Heavy metal monitoring in food-producing ecosystems can play an important role in human health safety. Since they are able to interfere with plants’ physiochemical characteristics, which influence the optical properties of leaves, they can be measured by in-field spectroscopy. In this study, the predictive power of spectroscopic data is examined. Five treatments of heavy metal stress (Cu, Zn, Pb, Cr, and Cd) were applied to grapevine seedlings and hyperspectral data (350−2500 nm), and heavy metal contents were collected based on in-field and laboratory experiments. The partial least squares (PLS) method was used as a feature selection technique, and multiple linear regressions (MLR) and support vector machine (SVM) regression methods were applied for modelling purposes. Based on the PLS results, the wavelengths in the vicinity of 2431, 809, 489, and 616 nm; 2032, 883, 665, 564, 688, and 437 nm; 1865, 728, 692, 683, and 356 nm; 863, 2044, 415, 652, 713, and 1036 nm; and 1373, 631, 744, and 438 nm were found most sensitive for the estimation of Cu, Zn, Pb, Cr, and Cd contents in the grapevine leaves, respectively. Therefore, visible and red-edge regions were found most suitable for estimating heavy metal contents in the present study. Heavy metals played a significant role in reforming the spectral pattern of stressed grapevine compared to healthy samples, meaning that in the best structures of the SVM regression models, the concentrations of Cu, Zn, Pb, Cr, and Cd were estimated with R2 rates of 0.56, 0.85, 0.71, 0.80, and 0.86 in the testing set, respectively. The results confirm the efficiency of in-field spectroscopy in estimating heavy metals content in grapevine foliage.

The quantitative structure–activity relationship (QSAR) analyses were carried out in a series of novel sulfonamide derivatives as the procollagen C-proteinase inhibitors for treatment of fibrotic conditions. Sphere exclusion method was used to classify data set into categories of train and test set at different radii ranging from 0.9 to 0.5. Multiple linear regression (MLR), principal component regression (PCR) and partial least squares (PLS) were used as the regression methods and stepwise, Genetic algorithm (GA), and simulated annealing (SA) were used as the feature selection methods. Three of the statistically best significant models were chosen from the results for discussion. Model 1 was obtained by MLR–SA methodology at a radius of 1.6. This model with a coefficient of determination ( r 2 ) = 0.71 can well predict the real inhibitor activities. Cross-validated q 2 of this model, 0.64, indicates good internal predictive power of the model. External validation of the model (pred_ r 2 = 0.85) showed that the model can well predict activity of novel PCP inhibitors. The model 2 which developed using PLS–SW explains 72% ( r 2 = 0.72) of the total variance in the training set as well as it has internal ( q 2 ) and external (pred_ r 2 ) predictive ability of ∼67% and ∼71% respectively. The last developed model by PCR–SA has a correlation coefficient ( r 2 ) of 0.68 which can explains 68% of the variance in the observed activity values. In this case internal and external validations are 0.61 and 0.75, respectively. Alignment Independent (AI) and atomic valence connectivity index (chiv) have the greatest effect on the biological activities. Developed models can be useful in designing and synthesis of effective and optimized novel PCP inhibitors which can be used for treatment of fibrotic conditions. [ABSTRACT FROM AUTHOR]

Protein kinases are essential components of various signaling pathways and represent attractive targets for therapeutic interventions. Kinase inhibitors are currently used to treat malignant tumors, as well as autoimmune diseases, due to their involvement in immune cell signaling. In this study, three-dimensional quantitative structure–activity relationship (3D-QSAR) analyses, including Multiple Linear Regression (MLR), Partial Least Squares (PLS), Multiple Non-Linear Regression (MNLR), Artificial Neural Network (ANN) and cross-validation analyses, were performed on a set of P38 MAP kinases as anti-cancer agents. This method, which is based on molecular modeling (molecular mechanics, Hartree-Fock (HF)), was used to determine the structural parameters, electronic properties, and energy associated with the molecules we examined. MLR, PLS, and MNLR analyses were performed on 46 protein P38 MAP kinase analogs to determine the relationships between molecular descriptors and the anti-cancer properties of the P38 MAP kinase analogs. The MLR model was validated by the external validation and standardization approach. The ANN, given the descriptors obtained from the MLR, exhibited a correlation coefficient close to 0.94. The predicted model was confirmed by two methods, leave-one-out (LOO) cross-validation and scrambling (or Y-randomization). We observed a high correlation between predicted and experimental activity, thereby both validating and demonstrating the high quality of the QSAR model that we described. [ABSTRACT FROM AUTHOR]

Quantitative structure–activity relationship (QSAR) models were proposed to correlate structural features or property descriptors of compounds with their corresponding biological activities. Because of the huge number of descriptors that encode different structural features used to generate valid QSAR models, variable selection becomes a fundamental step in building predictive and interpretative models. In this study, we applied a combined approach using multiple linear regression (MLR) and partial least-squares regression (PLS) to generate robust QSAR models with only a few descriptors applied to a set of cruzain inhibitors, namely, 61 semicarbazones and analogues, taken from the literature. From the 4885 descriptors generated by the Dragon program, we selected only five descriptors, applying the “Best-First” algorithm and PLS, followed by analysis of frequency and, finally, the genetic algorithm with MLR. The most significant QSAR equation encodes important steric and electronic structural features, which helps to identify in the set, structures that increase or decrease the pIC 50 values measured against cruzain. [ABSTRACT FROM AUTHOR]

Quantitative structure activity relationship were used to study a series of curcumin-related compounds with inhibitory effect on prostate cancer PC-3 cells, pancreas cancer Panc-1 cells, and colon cancer HT-29 cells. Sphere exclusion method was used to split data set in two categories of train and test set. Multiple linear regression, principal component regression and partial least squares were used as the regression methods. In other hand, to investigate the effect of feature selection methods, stepwise, Genetic algorithm, and simulated annealing were used. In two cases (PC-3 cells and Panc-1 cells), the best models were generated by a combination of multiple linear regression and stepwise (PC-3 cells: r 2 = 0.86, q 2 = 0.82, pred_r 2 = 0.93, and r 2 m (test) = 0.43, Panc-1 cells: r 2 = 0.85, q 2 = 0.80, pred_r 2 = 0.71, and r 2 m (test) = 0.68). For the HT-29 cells, principal component regression with stepwise (r 2 = 0.69, q 2 = 0.62, pred_r 2 = 0.54, and r 2 m (test) = 0.41) is the best method. The QSAR study reveals descriptors which have crucial role in the inhibitory property of curcumin-like compounds. 6ChainCount, T_C_C_1, and T_O_O_7 are the most important descriptors that have the greatest effect. With a specific end goal to design and optimization of novel efficient curcumin-related compounds it is useful to introduce heteroatoms such as nitrogen, oxygen, and sulfur atoms in the chemical structure (reduce the contribution of T_C_C_1 descriptor) and increase the contribution of 6ChainCount and T_O_O_7 descriptors. Models can be useful in the better design of some novel curcumin-related compounds that can be used in the treatment of prostate, pancreas, and colon cancers. [ABSTRACT FROM AUTHOR]

Herein we have studied the cytotoxicity and quantitative structure–activity relationship (QSAR) of heterocyclic compounds containing cyclic urea and thiourea nuclei. A set of 22 hydantoin and thiohydantoin related heterocyclic compounds were investigated with respect to their LC 50 values (Log of LC 50 ) against brine shrimp lethality bioassay in order to derive the 2D-QSAR models using MLR, PLS and ANN methods. The best predictive models by MLR, PLS and ANN methods gave highly significant square correlation coefficient ( R 2 ) values of 0.83, 0.81 and 0.91 respectively. The model also exhibited good predictive power confirmed by the high value of cross validated correlation coefficient Q 2 (0.74). [ABSTRACT FROM AUTHOR]

The ability to estimate chemical and physical properties from experimental spectra is highly desirable, as it eliminates the need for a priori knowledge of exact chemical structure and allows the property estimation of mixtures. Here we report the proof of principle that a predictive method for octanol-water partition coefficient (log P) based on 1H-NMR spectra in d3-chloroform is feasible and can yield accuracy comparable to in silico log P models. The Spectrometric Data-Activity Relationship (QSDAR) reported predicts log P of neutral organic chemicals using descriptors derived only from 1H-NMR chemical shifts, integrations and peak widths. Proton NMR spectra of 140 compounds with diverse structures were used to construct a Multiple Linear Regression (MLR) and a Partial Least Squares (PLS) model that predicts log P. The optimized models were internally validated by K-fold cross validation and leave-one-out validation, and externally with a test set of 28 chemicals. The squared regression coefficients of prediction for the MLR and PLS regression models were 0.970 and 0.971 respectively, showing that the method allows accurate prediction of log P values exclusively from predicted 1H NMR spectra. [ABSTRACT FROM AUTHOR]

Abstract: Banana fruit quality and maturity stages were studied at three different temperatures, viz., 20, 25, and 30°C by using hyperspectral imaging technique in the visible and near infrared (400–1000nm) regions. The quality parameters like moisture content, firmness and total soluble solids were determined and correlated with the spectral data. The spectral data were analyzed using the partial least square analysis. The optimal wavelengths were selected using predicted residual error sum of squares. The principal component analysis was also used to test the variability of the observed data. By using multiple linear regressions (MLR), models were established based on the optimal wave lengths to predict the quality attributes. The coefficient of determination was found to be 0.85, 0.87, and 0.91 for total soluble solids, moisture and firmness of the banana fruits, respectively. The change in TSS and firmness of banana fruits stored at different temperatures, viz., 20, 25, and 30°C during the ripening process followed the polynomial relationships and the change in moisture content followed a linear relationship at different maturity stages. [Copyright &y& Elsevier]

Quantitative structure activity relationship studies are indubitably of great importance in modern chemistry. In pursuit of better antihypertensive agents, QSAR studies were performed on a series of 48 N-substituted adenosines analogs. The models were developed using multiple linear regression (MLR) and partial least square (PLS) with many descriptors like electronic, topological, and lipophilic. QSAR models were evaluated for statistical significance and predictive power. The MLR and PLS generated comparable results with good predictive ability and all the other statistical values such as r, r, r, F, and s values were 0.92, 0.84, 0.83, 61.74, 0.24, respectively, for MLR and r and statistical significance values of 0.82 and 0.80, respectively, for PLS were satisfactory. The model indicates that the Verloop L, total lipole, VAMP nuclear energy, plays an important role in determining the antihypertensive activity of N-substituted adenosine receptor agonists. The model developed in the present study will be useful in the design of more potent N-substituted adenosine receptor agonists as antihypertensive agent. [ABSTRACT FROM AUTHOR]

Abstract : As part of an effort to establish a structure-activity relationship of diamidines against African trypanosomes, a quantitative correlation between molecular structure and anti-trypanosomal activity of 2-phenylbenzofuran derivatives was attained using classical quantitative structure-activity relationship (QSAR) descriptors and 3D similarity indices. A good model was obtained on the basis of classical descriptors; however, the model derived using descriptors based on similarity indices neither complemented the classical descriptors nor were significantly predictive. The best QSAR model with chemical descriptors that showed good correlative and predictive ability with r = 0.91, r = 0.82, and r = 0.80 was developed using stepwise multiple linear regression analysis (MLR) and a comparable partial least squares analysis (PLS) model with r = 0.79 was also obtained. The QSAR models revealed that a substituent steric descriptor (Verloop B parameter) and geometrical (moment of inertia 3 length) and hydrophobic (log P) descriptors of the whole molecule have significant impact on anti-trypanosomal activity of the compounds. The best QSAR models were validated by the leave one out technique. To further confirm the predictive power of the models, an external set of molecules was used which was not part of the training set. The high agreement betwPLS as shown in Eqeen experimental and predicted inhibitory values, obtained in the validation procedure, indicates the good quality of the derived QSAR models. In addition to QSAR analysis Lipinski's rule was also applied to the series under consideration and newly designed molecules in order to check the drugability of the compounds; no violation of this rule was found. Hence 2-phenylbenzofuran has tremendous potential to yield orally active anti-trypanosomal agents. Graphical abstract: [Figure not available: see fulltext.] [ABSTRACT FROM AUTHOR]

The areas of signals present in 12 consecutive frequency bins of the 1H nuclear magnetic resonance (NMR) spectra of different vacuum residues (VRs) were correlated with the yield of the deasphalting/demetallization process (DEMEX) under fixed and variable conditions (antisolvent/load ratio and rectification temperature). Prediction models for the demetallized oil (DMO) obtained under variable process conditions generated by partial least squares (PLS) regression had R2 = 0.950 and the CV-q2 was 0.943. Models obtained with multiple linear regression (MLR) for the same process had R2 = 0.93 but with a significantly lower CV-q2 (0.608) value. The prediction models for the DMO yield under a set of fixed conditions was such that R2 = 0.930 for the PLS whereas for the MRL method it was 0.970. Again, the CV-q2 value was higher for the PLS (0.998) than for the MRL method (0.610). The proposed correlation methods are much faster and significantly less costly than the conventional ones used for predicting the yield of the DMO process. [ABSTRACT FROM AUTHOR]

QSAR analysis using multiple linear regression and partial least squares methods were conducted on a data set of 47 pyrrolidine analogs acting as DPP IV inhibitors. The QSAR models generated (both MLR and PLS) were robust with statistically significant s, F, r, r2 and r2 (CV) values. The analysis helped to ascertain the role of shape flexibility index, Ipso atom E-state index and electrostatic parameters like dipole moment, in determining the activity of DPP IV inhibitors. In addition to QSAR modeling, Lipinski''s rule of five was also employed to check the pharmacokinetic profile of DPP IV inhibitors. Since none of the compounds violated the Lipinski''s rule of five indicating that the DPP IV inhibitors reported herein have sound pharmacokinetic profile and can be considered as potential drug candidates for diabetes mellitus Type II. [ABSTRACT FROM AUTHOR]

The quantitative structure activity relationship (QSAR) models were developed using multiple linear regression (MLR) and partial least square (PLS) for a set of 85 AT1 receptor antagonists of hydantoin series. The MLR and PLS generated comparable models with good predictive ability and all the other statistical values, such as r, r2, $$ {\text{r}}_{{({\text{cv}})}}^{2} , $$ and F and s values, were satisfactory. The results obtained from this study indicate the importance of steric (K-alpha3), hydrophobic (log P, and total lipole), and total energy (Cosmic total energy) in determining the activity of AT1 receptor antagonists. The results clearly explained that optimum hydrophobicity of substituent at R2 position is favorable for the activity and presence of a substituent of particular size and shape on phenyl ring at R3 position is essential for the activity. This information is pertinent to the further design of new AT1 receptor antagonist containing the hydantoin nucleus. [ABSTRACT FROM AUTHOR]

In this work, some chemometrics methods were applied for modelling and prediction of the acidic dissociation constants of some organic compounds with descriptors calculated from the molecular structure alone. The stepwise multiple linear regression method was used to select descriptors which are responsible for the pKa of these compounds. Then support vector machine (SVM), principal component regression (PCR), partial least squares (PLS) and multiple linear regression (MLR) were utilized to construct the nonlinear and linear quantitative structure-activity relationship models. The results obtained using SVM were compared with PLS, PCR and MLR, revealing that the SVM model was much better than other models. The root-mean-square errors of the training set and the test set for the SVM model are 0.2551 and 0.6139, and the correlation coefficients were 0.9936 and 0.9919, respectively. This paper provides a new and effective method for predicting pKa of organic compounds, and also reveals that SVM can be used as a powerful chemometrics tool for QSPR studies. Finally, results have shown that the SVM drastically enhances the ability of prediction in QSAR studies superior to multiple linear regression, principal component regression and partial least squares. [ABSTRACT FROM AUTHOR]

Abstract: Quantitative structure–retention relationships have been developed to predict retention indices of fatty acid methyl esters on standard non-polar polydimethylsiloxane stationary phases. Branched, saturated and unsaturated compounds were included. All retention indices have been evaluated by statistical processing of experimentally measured and literature data in accordance with the concept of interlaboratory data randomization. Multiple linear regression (MLR) has been carried out to find relationships between selected properties and retention indices. Models have been built in two different ways (i) the same degrees of freedom for all models have been fixed and the variable selection ability has been compared; (ii) variable selection methods have been used in their best performance. The five selection methods were: pair-wise correlation, forward selection, partial least squares projection of latent structures, modified best subset selection and the Lasso method. The stability and the validity of models have been tested by internal and external validation. The error of predicted retention indices is close to the error of interlaboratory reproducibility of retention indices. The most relevant variables in description of retention indices were molecular mass, number of double bonds and number of rotatable bonds complemented with topological descriptors. Predictive models have been built for 130 fatty acid methyl esters for identification purposes. Moreover, prediction of unknown retention indices for 37 fatty acid methyl esters has also been carried out. [Copyright &y& Elsevier]

Abstract: A quantitative structure–property relationship (QSPR) study is suggested for the prediction of toxicity (IGC50) of nitrobenzenes. Ab initio theory was used to calculate some quantum chemical descriptors including electrostatic potentials and local charges at each atom, HOMO and LUMO energies, etc. Modeling of the IGC50 of nitrobenzenes as a function of molecular structures was established by means of the least squares support vector machines (LS-SVM). This model was applied for the prediction of the toxicity (IGC50) of nitrobenzenes, which were not in the modeling procedure. The resulted model showed high prediction ability with root mean square error of prediction of 0.0049 for LS-SVM. Results have shown that the introduction of LS-SVM for quantum chemical descriptors drastically enhances the ability of prediction in QSAR studies superior to multiple linear regression and partial least squares. [Copyright &y& Elsevier]

Retention indices for a new series of methyl‐thiazole‐oxadiazolines were determined by reversed phase high performance thin layer chromatography on C 18 plates with methanol‐water in different volume proportions as mobile phase. Comparisons based on the multiple regression methods including multiple linear regression (MLR), principal component regression (PCR) and partial least squares (PLS) have been applied to the modelling of chromatographic lipophilicity (R Mo and b values) by means of 16 different descriptors obtained by using Alchemy package software. The results achieved concerning the prediction of retention indices are highly significant and are in very good agreement with the molecular structure of the compounds investigated. The largest positive charge over the atoms in a molecule (Q + ), the sum of absolute values of the charges on each atom of the molecule (SQ), and the sum of absolute values of the charges on the nitrogens and oxygens in the molecule (SQ NO ) seem to be dominant in the retention mechanism. [ABSTRACT FROM AUTHOR]

Abstract: Attempts have been made to assess previous heat treatment of fish-meat gels prepared from walleye pollack and horse mackerel surimi, since surimi-based products have been gaining popularity for their protein quality, law fat content and convenience in consumption. Visible–NIR spectra of gels (30–90°C) were collected from 650 to 1100nm with a surface interactance fibre optic accessory. Partial least squares (PLS) and multiple linear regression (MLR) techniques were employed for data analysis. Spectral changes upon heat treatment were related to the heating temperature which reflected the changes in the environment of the secondary structure due to the denaturation of proteins, and to changes in the state of water. A promising linear relationship (R =0.98) was observed between NIR-predicted temperatures and the actual heating temperatures with prediction error of 1.85°C. [Copyright &y& Elsevier]

Abstract: The term bagging refers to a class of techniques in which an ensemble model is obtained by combining different member models generated by resampling the available data set. It has been shown that bagging can lead to substantial gains in accuracy for both classification and regression models, specially when alterations in the training set cause significant changes in the outcome of the modelling procedure. However, in the context of chemometrics, the use of bagging for quantitative multicomponent analysis is still incipient. More recently, an alternative aggregation scheme termed subagging, which is based on subsampling without replacement, has been shown to provide performance improvements similar to bagging at a smaller computational cost. The present paper proposes a strategy for using subagging in conjunction with three multivariate calibration methods, namely Partial Least Squares (PLS) and Multiple Linear Regression with variable selection by using either the Successive Projections Algorithm (MLR-SPA) or a Genetic Algorithm (MLR-GA). The subagging member models are generated by subsampling the pool of samples available for modelling and then forming new calibration sets. Such a strategy is of value in analytical problems involving complex matrices, in which reproducing the composition variability of real samples by means of optimized experimental designs may be a difficult task. The efficiency of the proposed strategy is illustrated in a problem involving the NIR spectrometric determination of four diesel quality parameters (specific mass, sulphur content, and the distillation temperatures T10% and T90% at which 10% and 90% of the sample has evaporated, respectively). In this case study, the use of 30 subsampling iterations provides relative improvements of up to 16%, 33%, and 35% in the prediction accuracy of PLS, MLR-SPA, and MLR-GA models, respectively, with respect to the expected results of individual (non-ensemble) models. [Copyright &y& Elsevier]

Abstract: Noninvasive near-infrared (NIR) spectroscopy was employed to assess the endpoint temperatures (EPT) of kamaboko gel (n=140) prepared at different temperatures. Partial Least-Squares (PLS) and Multiple Linear Regressions (MLR) were used to develop a calibration which was tested with a validation set. Spectral changes upon heat treatment were related to the heating temperature which might be the reason for changes in the secondary structure of the gels due to the denaturation of proteins, and to changes in the state of water. The changes are most clearly seen in the 1300–1600nm region in bands mainly attributed to absorption by functional groups in protein. NIR-predicted EPT and actual heating temperatures revealed a linear relationship (R=0.98). Between 30 and 90°C, NIR spectroscopy detected EPT of kamaboko gels with a prediction error of 2.16°C. The MLR and PLS calibration had similar performance for determining EPT of kamaboko gel if the appropriate wavelengths or wavelength region were selected. [Copyright &y& Elsevier]

In this study, quantitative structure–retention relationship (QSRR) technique was used to find the best approximation and to predict gas chromatographic retention indices for O-, N-, and S-heterocyclic compounds on standard nonpolar polydimethyl siloxane stationary phase. Boiling point (BP) and calculated properties were used to encode the structure of compounds. Three- and two-dimensional calculated properties such as weighted–holistic invariant molecular (WHIM) descriptors, geometry topology and atom weights assembly (GETAWAY) descriptors, connectivity indices, and zero-dimensional constitutive descriptors were used. Variable subset selection (VSS) and partial least squares (PLS) projections to latent structures were used to select the most significant variables from a large set of descriptors. Multiple linear regression (MLR) and PLS were applied to find the relationship between selected properties and gas chromatographic retention indices. PLS was not able to select the most important descriptors (boiling point or molecular weight). The predictive ability of the models was tested by cross-validation. Solely calculated descriptors were not able to give proper models. Boiling point was always necessary for good prediction. PLS models containing boiling points were suitable for retention index prediction, whereas MLR did not give real linear models. [Copyright &y& Elsevier]

A quantitative structure–retention relationship (QSRR) study has been performed for the correlation of the gas chromatographic-mass spectrometry (GC-MS) relative retention times (RRT) of α-, β1-, and β2-agonists with their molecular characteristics in order to create models for the prediction of RRT values of unanalyzed molecules. The data structure was investigated by principal component analysis (PCA). Modeling of the RRT of α-, β1-, and β2-agonists, as a function of many diverse descriptors, was established by means of the multiple linear regression (MLR) and partial least-squares projections to latent structures (PLS) methods. [Copyright &y& Elsevier]

The variable selection in QSAR studies by MLR and PLS modeling has been performed using the evolution algorithm (EA). The Cp statistic has been modified and used as the objective function in the EA search for different combinations of molecular descriptors. For MLR modeling a few information-rich descriptors are selected for model formulation. In PLS modeling, the proposed procedure selects a relatively large number of information-containing descriptors, and a PLS model is formulated based on a few latent variables, which are linear combinations of the selected descriptors. The proposed procedures were used for the prediction of carcinogenicity of aromatic amines. [ABSTRACT FROM AUTHOR]

Quantitative structure–activity relationship (QSAR) analysis has been directed to a series of recently synthesized nifedipine analogues containing the nitroimidazolyl group at the C-4 position and different ester substituents at C-3 and C-5 positions of the 1,4-dihydropyridine (DHP) ring. Modeling of the calcium channel antagonist activity of these compounds as a function of the theoretically derived descriptors was established by multiple linear regression (MLR) and partial least squares (PLS) regression. The genetic algorithm (GA) was used for the selection of the variables that resulted in the best-fitted models. Appropriate models with low standard errors and high correlation coefficients were obtained. Topology, hydrophobicity and surface area examples of the descriptors affected by the binding of dihydropyridine to the receptor. A comparison of the two regression methods used showed that PLS has a better prediction ability than MLR. [Copyright &y& Elsevier]

Table grape quality is of importance for consumers and thus for producers. Its objective quality is usually determined by destructive methods mainly based on sugar content. This study proposed to evaluate the possibility of hyperspectral imaging to characterize table grapes quality through its sugar (TSS), total flavonoid (TF), and total anthocyanin (TA) contents. Different data pre-treatments (WD, SNV, and 1st and 2nd derivative) and different methods were tested to get the best prediction models: PLS with full spectra and then Multiple Linear Regression (MLR) were realized after selecting the optimal wavelengths thanks to the regression coefficients (β-coefficients) and the Variable Importance in Projection (VIP) scores. All models were good at showing that hyperspectral imaging is a relevant method to predict sugar, total flavonoid, and total anthocyanin contents. The best predictions were obtained from optimal wavelength selection based on β-coefficients for TSS and from VIPs optimal wavelength windows using SNV pre-treatment for total flavonoid and total anthocyanin content. Thus, good prediction models were proposed in order to characterize grapes while reducing the data sets and limit the data storage to enable an industrial use. [ABSTRACT FROM AUTHOR]