Beilei Lei

QSAR study of malonyl‐CoA decarboxylase inhibitors using GA‐MLR and a new strategy of consensus modeling

Quantitative structure‐activity relationship (QSAR) of a series of structural diverse malonyl‐CoA decarboxylase (MCD) inhibitors have been investigated by using the predictive single model as well as the consensus analysis based on a new strategy proposed by us. Self‐organizing map (SOM) neural network was employed to divide the whole data set into representative training set and test set. Then a multiple linear regressions (MLR) model population was built based on the theoretical molecular descriptors selected by Genetic Algorithm using the training set. In order to analyze the diversity of these models, multidimensional scaling (MDS) was employed to explore the model space based on the Hamming distance matrix calculated from each two models. In this space, Q2 (cross‐validated R2) guided model selection (QGMS) strategy was performed to select submodels. Then consensus modeling was built by two strategies, average consensus model (ACM) and weighted consensus model (WCM), where each submodel had a different weight according to the contribution of model expressed by MLR regression coefficients. The obtained results prove that QGMS is a reliable and practical method to guide the submodel selection in consensus modeling building and our weighted consensus model (WCM) strategy is superior to the simple ACM.

Molecular modeling studies of vascular endothelial growth factor receptor tyrosine kinase inhibitors using QSAR and docking.

The vascular endothelial growth factor (VEGF) and its receptor tyrosine kinases VEGFR-2 or kinase insert domain receptor (KDR) are attractive targets for the development of novel anticancer agents. In the present work, comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) were performed on a series of selective inhibitors of KDR. Docking studies were performed to explore the binding mode between all of the inhibitors and the KDR and produce the bioactive conformation of each compound in the whole dataset. Two conformer-based alignment strategies were employed to construct reliable 3D-QSAR models. The docked conformer-based alignment strategy gave the best 3D-QSAR models. The best CoMFA and CoMSIA models gave a cross-validated coefficient q(2) of 0.546 and 0.715, non-cross-validated r(2) values of 0.936 and 0.961, predicted r(2) values of 0.673 and 0.797, respectively. The information obtained from molecular modeling studies were very helpful to design some novel selective inhibitors of KDR with desired activity.

Global, local and novel consensus quantitative structure-activity relationship studies of 4-(Phenylaminomethylene) isoquinoline-1, 3 (2H, 4H)-diones as potent inhibitors of the cyclin-dependent kinase 4

Quantitative structure-activity relationship (QSAR) studies on a series of selective inhibitors of the cyclin-dependent kinase 4 (CDK4) were performed by using two conventional global modeling methods (multiple linear regression (MLR) and support vector machine (SVM)), local lazy regression (LLR) as well as three consensus models. It is remarkable that the LLR model could improve the performance of the QSAR model significantly. In addition, due to the fact that each model can predict certain compounds more accurately than other models, the above three derived models were all used as submodels to build consensus models using three different strategies: average consensus model (ACM), simple weighted consensus model (SWCM) and hat weighted consensus model (HWCM). Through the analysis of the results, the HWCM consensus strategy, firstly proposed in this work, proved to be more reliable and robust than the best single LLR model, ACM and SWCM models.

A novel method for protein‐ligand binding affinity prediction and the related descriptors exploration

In this study, a novel method was developed to predict the binding affinity of protein‐ligand based on a comprehensive set of structurally diverse protein‐ligand complexes (PLCs). The 1300 PLCs with binding affinity (493 complexes with Kd and 807 complexes with Ki) from the refined dataset of PDBbind Database (release 2007) were studied in the predictive model development. In this method, each complex was described using calculated descriptors from three blocks: protein sequence, ligand structure, and binding pocket. Thereafter, the PLCs data were rationally split into representative training and test sets by full consideration of the validation of the models. The molecular descriptors relevant to the binding affinity were selected using the ReliefF method combined with least squares support vector machines (LS‐SVMs) modeling method based on the training data set. Two final optimized LS‐SVMs models were developed using the selected descriptors to predict the binding affinities of Kd and Ki. The correlation coefficients (R) of training set and test set for Kd model were 0.890 and 0.833. The corresponding correlation coefficients for the Ki model were 0.922 and 0.742, respectively. The prediction method proposed in this work can give better generalization ability than other recently published methods and can be used as an alternative fast filter in the virtual screening of large chemical database.

Novel approaches for retention time prediction of oligonucleotides in ion-pair reversed-phase high-performance liquid chromatography

The base sequence autocorrelation (BSA) descriptors were used to describe structures of oligonucleotides and to develop accurate quantitative structure-retention relationship (QSRR) models of oligonucleotides in ion-pair reversed-phase high-performance liquid chromatography. Through the combination use of multiple linear regression (MLR) and genetic algorithm (GA), QSRR models were developed at temperatures of 30 degrees C, 40 degrees C, 50 degrees C, 60 degrees C and 80 degrees C, respectively. Satisfactory results were obtained for the single-temperature models (STM). Multi-temperature model (MTM) was also developed that can be used for predicting the retention time at any temperature. The correlation coefficients of retention time prediction for the test set based on the MTM model at 30 degrees C, 40 degrees C, 50 degrees C, 60 degrees C and 80 degrees C were 0.978, 0.982, 0.989, 0.988 and 0.996, respectively. The corresponding absolute average relative deviations (AARD) for the test set at each temperature were all less than 1%. The new strategy of feature representation and multi-temperatures modeling is a very promising tool for QSRR modeling with good predictive ability for predicting retention time of oligonucleotides at multiple temperatures under the studied condition.

Rational prediction of the herbicidal activities of novel protoporphyrinogen oxidase inhibitors by quantitative structure-activity relationship model based on docking-guided active conformation.

Molecular docking-guided active conformation selection was used in a quantitative structure-activity relationship (QSAR) study of a series of novel protoporphyrinogen oxidase (PPO) inhibitors with herbicidal activities. The developed model can be used for the rational and accurate prediction of herbicidal activities of these inhibitors from their molecular structures. Molecular docking study was carried out to dock the inhibitors into the PPO active site and to obtain the rational active conformations. Based on the conformations generated from molecular docking, satisfactory predictive results were obtained by a genetic algorithm-multiple linear regression (GA-MLR) model according to the internal and external validations. The model gave a correlation coefficient R(2) of 0.972 and 0.953 and an absolute average relative deviation AARD of 2.24% and 2.75% for the training set and test set, respectively. The results from this work demonstrate that the molecular docking-guided active conformation selection strategy is rational and useful in the QSAR study of these PPO inhibitors and for the quantitative prediction of their herbicidal activities. The results obtained could be helpful to the design of new derivatives with potential herbicidal activities.

Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) of thiazolone derivatives as hepatitis C virus NS5B polymerase allosteric inhibitors

Three-dimensional quantitative structure-activity relationship (3D-QSAR) models for a series of thiazolone derivatives as novel inhibitors bound to the allosteric site of hepatitis C virus (HCV) NS5B polymerase were developed based on CoMFA and CoMSIA analyses. Two different conformations of the template molecule and the combinations of different CoMSIA field/fields were considered to build predictive CoMFA and CoMSIA models. The CoMFA and CoMSIA models with best predictive ability were obtained by the use of the template conformation from X-ray crystal structures. The best CoMFA and CoMSIA models gave q2 values of 0.621 and 0.685, and r2 values of 0.950 and 0.940, respectively for the 51 compounds in the training set. The predictive ability of the two models was also validated by using a test set of 16 compounds which gave rpred2 values of 0.685 and 0.822, respectively. The information obtained from the CoMFA and CoMSIA 3D contour maps enables the interpretation of their structure-activity relationship and was also used to the design of several new inhibitors with improved activity.

Molecular modeling studies of Rho kinase inhibitors using molecular docking and 3D-QSAR analysis

Rho kinase (ROCK) has become an attractive target for the treatment of many diseases such as hypertension, stroke and cancer. In this work, molecular docking and three-dimensional quantitative structure-activity relationship (3D-QSAR) studies were performed on a series of ROCK inhibitors. Molecular docking was used to explore the binding mode between the ligands and the receptor. Based on the docked conformations, comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) were performed to gain insights into the key structural factors affecting the activity. The results of the molecular modeling studies suggested that further modification of these inhibitors with bulky and hydrophobic groups that accommodated in the distal region of the ROCK binding pocket would improve the activity.

Global and local prediction of protein folding rates based on sequence autocorrelation information.

To understand the folding behavior of proteins is an important and challenging problem in modern molecular biology. In the present investigation, a large number of features representing protein sequences were developed based on sequence autocorrelation weighted by properties of amino acid residues. Genetic algorithm (GA) combined with multiple linear regression (MLR) was employed to select significant features related to protein folding rates, and to build global predictive model. Moreover, local lazy regression (LLR) method was also used to predict the protein folding rates. The obtained results indicated that LLR performed much better than the global MLR model. The important properties of amino acid residues affecting protein folding rates were also analyzed. The results of this study will be helpful to understand the mechanism of protein folding. Our results also demonstrate that the features of amino acid sequence autocorrelation is effective in representing the relationship between protein sequence and folding rates, and the local method is a powerful tool to predict the protein folding rates.

Structure‐based quantitative structure‐activity relationship studies of checkpoint kinase 1 inhibitors

Structure‐based quantitative structure‐activity relationship (QSAR) studies on a series of checkpoint kinase 1 (Chk1) inhibitors were performed to find the key structural features responsible for their inhibitory activity. Molecular docking was employed to explore the binding mode of all inhibitors at the active site of Chk1 and determine the active conformation for the QSAR studies. Ligand and structure‐based descriptors incorporating the ligand‐receptor interaction were generated based on the docked complex. Genetic Algorithm‐Multiple Linear Regression (GA‐MLR) method was used to build 2D QSAR model. The 2D QSAR model gave a squared correlation coefficient R2 of 0.887, cross‐validated Q2 of 0.837 and the prediction squared correlation coefficient R  2pred of 0.849, respectively. Furthermore, three‐dimensional quantitative structure‐activity relationship (3D QSAR) model using comparative molecular field analysis (CoMFA) with R2 of 0.983, Q2 of 0.550 and R  2pred of 0.720 was also developed. The obtained results are helpful for the design of novel Chk1 inhibitors with improved activities.