QSAR study of heparanase inhibitors activity using artificial neural networks and Levenberg–Marquardt algorithm

Jalali‐Heravi, Mehdi; Asadollahi-Baboli, M.; Shahbazikhah, Parviz

doi:10.1016/j.ejmech.2007.04.014

Cited by 84 publications

(35 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In this study we used levenberg-marquardt algorithm for training the network. The detailed discussion about training algorithms for ANN can be found elsewhere [21].…”

Section: Artificial Neural Networkmentioning

confidence: 99%

QSAR Modeling of 1‐(3,3‐Diphenylpropyl)‐Piperidinyl Amides as CCR5 Modulators Using Multivariate Adaptive Regression Spline and Bayesian Regularized Genetic Neural Networks

Jalali‐Heravi

Mani-Varnosfaderani

2009

QSAR Comb. Sci.

View full text Add to dashboard Cite

This study deals with developing a quantitative structure-activity relationship (QSAR) model for describing and predicting the inhibition activity of 1-(3,3-diphenylpropyl)-piperidinyl derivatives as CCR5 modulators. Applying the multiple linear regressions (MLR) and its inability in predicting the inhibition behavior showed that the interaction has no linear characteristics. To assess the nonlinear characteristics of the inhibition activity artificial neural networks (ANN) was used for data modeling. In order to select the variables needed for developing ANNs, three variable selection algorithms were used: Stepwise-MLR, genetic algorithm-partial least squares (GA-PLS), and Bayesian regularized genetic neural networks (BRGNNs). R 2 and root mean square error (RMSE) values for training (t) and leave-one-out (LOO) procedures revealed that BRGNNs is a robust algorithm for the variable selection and regression method simultaneously. Due to the black box limitation of neural networks, multivariate adaptive regression spline (MARS) technique was used for modeling. A prominent advantage of MARS with respect to ANN is its ability in interpreting of the results of the model. Q 2 LOO and R t 2 (0.982 and 0.947) reveal that MARS can describe and predict inhibition activity of these modulators and is as robust as ANN. Because the MARS model can explain the activity of molecules, it is a useful model for designing novel CCR5 inhibitors.

show abstract

“…In this study we used levenberg-marquardt algorithm for training the network. The detailed discussion about training algorithms for ANN can be found elsewhere [21].…”

Section: Artificial Neural Networkmentioning

confidence: 99%

QSAR Modeling of 1‐(3,3‐Diphenylpropyl)‐Piperidinyl Amides as CCR5 Modulators Using Multivariate Adaptive Regression Spline and Bayesian Regularized Genetic Neural Networks

Jalali‐Heravi

Mani-Varnosfaderani

2009

QSAR Comb. Sci.

View full text Add to dashboard Cite

show abstract

“…For the purpose of comparison, stepwise MLR method was also used for the variable selection and the selected descriptors were used as input of the ANN model for constructing the QSAR model (stepwise MLR-ANN) [18]. The theory of stepwise MLR-ANN method is given in our pervious work [11] and it is not discussed here for the sake of brevity. Having more acceptable R 2 and RMSE values for the calibration and validation sets, GA-ANN method is superior to the stepwise MLR-ANN method (see Tables 1 and 3).…”

Section: Ga-ann Strategy For Developing Qsar Modelmentioning

confidence: 99%

QSAR Analysis of Platelet‐derived Growth Inhibitors Using GA‐ANN and Shuffling Crossvalidation

Jalali‐Heravi

Asadollahi-Baboli

2008

QSAR Comb. Sci.

View full text Add to dashboard Cite

Quantitative Structure -Activity Relationship (QSAR) models for the inhibition action of some 1-phenylbenzimidazoles on platelet-derived growth are constructed using Genetic Algorithm and Artificial Neural Network (GA-ANN) method. The statistical parameters of R 2 and root-mean-square error are 0.82 and 0.21, respectively using this method. These parameters show a considerable improvement compared to the stepwise multiple linear regression combined with ANN (stepwise MLR-ANN). Ten-fold shuffling crossvalidations are carried out to select the most important descriptors. Five descriptors of index of Balaban (J), average molecular weight (AMW), 3D-Wiener index (W3D), mean atomic van der Waals volume (Sv), and total charge (Q total ) appear in most of the models. The results of GA-ANN are superior compared with those of GA-MLR and GA-PLS, which indicate that the inhibition behavior has nonlinear characteristics. The ability of GA-ANN model in predicting inhibition behavior of 1-phenylbenzimidazoles [log(1/IC 50 )] and its robustness is illustrated by validation techniques of leave-one-out and leave-multipleout crossvalidations and also by Y-randomization technique.

show abstract

“…[16] Therefore, the process of data mining is of great importance and many efforts have been made to develop different feature selection or extraction algorithms. [17][18][19][20] These algorithms are basically divided into three categories: filter, wrapper and embedded. [21] The first class, called filter models, select variables independently of the classifying method.…”

Section: Introductionmentioning

confidence: 99%

Toward an Optimal Approach for Variable Selection in Counter‐Propagation Neural Networks: Modeling Protein‐Tyrosine Kinase Inhibitory of Flavanoids Using Substituent Electronic Descriptors

Hemmateenejad

Mehdipour

Deeb

et al. 2011

Molecular Informatics

View full text Add to dashboard Cite

Counter propagation neural network (CPNN) is one of the attractive tools of classification in QSAR studies. A major obstacle in classification by CPNN is finding the best subset of variables. In this study, the performance of some different feature selection algorithms including F score-based ranking, eigenvalue ranking of PCs obtained from data set, Non-Error-Rate (NER) ranking of both descriptors and PCs, and 3-way handling of data, Parallel Factor Analysis (PARAFAC), was evaluated in order to find the best classification model. The methods were applied for modeling protein-tyrosine kinase inhibitory of some flavonoid derivatives using substituent electronic descriptors (SED) as novel source of electronic descriptors. The results showed that the best performance was achieved by F-score ranking while the NER ranking of principal components (PCs) showed very fluctuate results and the worst performance was belonging to PARAFAC-CPNN. Furthermore, comparison of results of these nonlinear algorithms with linear discriminate analysis method revealed better predictions by the former.

show abstract

QSAR study of heparanase inhibitors activity using artificial neural networks and Levenberg–Marquardt algorithm

Cited by 84 publications

References 32 publications

QSAR Modeling of 1‐(3,3‐Diphenylpropyl)‐Piperidinyl Amides as CCR5 Modulators Using Multivariate Adaptive Regression Spline and Bayesian Regularized Genetic Neural Networks

QSAR Modeling of 1‐(3,3‐Diphenylpropyl)‐Piperidinyl Amides as CCR5 Modulators Using Multivariate Adaptive Regression Spline and Bayesian Regularized Genetic Neural Networks

QSAR Analysis of Platelet‐derived Growth Inhibitors Using GA‐ANN and Shuffling Crossvalidation

Toward an Optimal Approach for Variable Selection in Counter‐Propagation Neural Networks: Modeling Protein‐Tyrosine Kinase Inhibitory of Flavanoids Using Substituent Electronic Descriptors

Contact Info

Product

Resources

About