Advanced
Designing Hypothesis of 2-Substituted-N-[4-(1-methyl-4,5-diphenyl-1H-imidazole-2-yl)phenyl] Acetamide Analogs as Anticancer Agents: QSAR Approach
Designing Hypothesis of 2-Substituted-N-[4-(1-methyl-4,5-diphenyl-1H-imidazole-2-yl)phenyl] Acetamide Analogs as Anticancer Agents: QSAR Approach
Journal of the Korean Chemical Society. 2013. Dec, 57(6): 744-754
Copyright © 2013, Korea Chemical Society
  • Received : June 02, 2013
  • Accepted : October 13, 2013
  • Published : December 20, 2013
Download
PDF
e-PUB
PubReader
PPT
Export by style
Article
Author
Metrics
Cited by
TagCloud
About the Authors
Ajay B. Bedadurge
Anwar R. Shaikh

Abstract
Quantitative structure–activity relationship (QSAR) analysis for recently synthesized imidazole-(benz)azole and imidazole - piperazine derivatives was studied for their anticancer activities against breast (MCF-7) cell lines. The statistically significant 2D-QSAR models (r 2 = 0.8901; q 2 = 0.8130; F test = 36.4635; r 2 se = 0.1696; q 2 se = 0.12212; pred_r 2 = 0.4229; pred_r 2 se = 0.4606 and r 2 = 0.8763; q 2 = 0.7617; F test = 31.8737; r 2 se = 0.1951; q 2 se = 0.2708; pred_r 2 = 0.4386; pred_r 2 se = 0.3950) were developed using molecular design suite (VLifeMDS 4.2). The study was performed with 18 compounds (data set) using random selection and manual selection methods used for the division of the data set into training and test set. Multiple linear regression (MLR) methodology with stepwise (SW) forward-backward variable selection method was used for building the QSAR models. The results of the 2D-QSAR models were further compared with 3D-QSAR models generated by kNN-MFA, (k-Nearest Neighbor Molecular Field Analysis) investigating the substitutional requirements for the favorable anticancer activity. The results derived may be useful in further designing novel imidazole-(benz)azole and imidazole-piperazine derivatives against breast (MCF-7) cell lines prior to synthesis.
Keywords
INTRODUCTION
DNA intercalators are the chemotherapeutics which exhibit anticancer activity by inserting between the base pairs of the double helix and causing a significant change of DNA conformation. 1 They bind to DNA by non-covalent interactions and constitute DNA-intercalator complex. The only recognized forces that maintain the stability of the DNA-intercalator complex, even more than DNA alone, are van der Waals, hydrogen bonding, polarization and hydrophobic forces. 2 6 The compounds which bear heteroatoms such as nitrogen, sulphur and oxygen increase the strength of the complex by forming hydrogen bonds with DNA. The force of interaction between compound and DNA usually correlates with the anticancer activity. 7 9 Besides, when one or more nitrogen heteroatoms exist on the chemical structure, intercalating chromophore possesses a polarized character and optimal interaction occurs. 6 Based on such reasons the presence of heteroatoms in the compounds plays an important role on exhibiting the anticancer activity.
Imidazole is a nitrogen containing heterocyclic ring which possesses biological and pharmaceutical importance. Thus, imidazole compounds have been an interesting source for researchers for more than a century. 10 Lepidiline A and B are the imidazole compounds which exhibit cytotoxicity against various types of human cancer cell lines at micromolar concentrations. 11 Dacarbazine, 12 zoledronic acid, 13 tipifarnib 14 15 and azathioprine 16 are the imidazole ring bearing anticancer agents. In addition to imidazole, some other azole and benz azole ring systems such as benzoxazole, benzothiazole, triazole, tetrazole, thiadiazole and thiazoline can be shown as pharmacophore groups which are responsible for the anticancer activity. 17 22 Resistance improvement against the former anticancer agents creates a research area in development of new anticancer agents. Nevertheless, it is rather difficult to develop a new agent which can selectively inhibit the proliferation of abnormal cells with least or no effect on normal cells. 23 Therefore, cancer chemotherapy is very important for medicinal chemists and the studies are still being carried onto develop new chemotherapeutic agents that are probable to indicate activity on various cancer types. Prompted from the observations described above and the chemotherapeutic value of the nitrogenous chemical scaffolds, synthesis and anticancer activity evaluation of some novel imidazole-(benz)azole and imidazole piperazine derivatives were examined in this study.
Traditional computer-assisted quantitative structure-activity relationship (QSAR) studies pioneered by C. Hansch et al. 1962 24 have been proved to be one of the useful approaches for accelerating the drug design process which help to correlate the bioactivity of compounds with structural descriptors. Recently synthesized imidazole-(benz)azole and imidazole - piperazine derivatives against the breast (MCF-7) carcinoma cell lines. To gain further insights into the structure- activity relationships of these derivatives and understand the mechanism of their substitutional specificity, we have performed 2D and 3D-QSAR on imidazole-(benz)azole and imidazole - piperazine derivatives using Multiple linear regression (MLR) and k-Nearest Neighbor Molecular Field Analysis (kNN-MFA), respectively. The significance of the QSAR models was evaluated using cross-validation tests, randomization tests and external test set prediction. The robust 2D/3D-models may be useful in further designing new candidates as potential anticancer activity imidazole-( benz) azole and imidazole - piperazine against breast (MCF-7) carcinoma cell lines of prior to synthesis.
MATERIAL AND METHOD
- Selection of Molecules
Data set of 18 imidazole-(benz)azole and imidazole - piperazine derivatives ( 1 ) collected from published literature 25 were taken for the present study. The affinity data of inhibitory activities were converted into IC 50 values to get the linear relationship in equation using the following formula: p IC 50 = −logIC 50 , where IC 50 value represents inhibitory activity in IC 50 (μM) ( 2 ). Molecules were rationally divided into the training set and test set based on the suggestions given by Alexander Tropsha et al. 26
PPT Slide
Lager Image
- Molecular Modeling
All computational experiments were performed using on Lenovo computer having genuine Intel Pentium i3 Core Processor and Windows XP operating system using the software Molecular Design Suite (VlifeMDS 4.2). 27 Structures were drawn using the 2D draw application and converted to 3D structures and subjected to an energy minimization and geometry optimization using Merck Molecular Force Field, force field and charges followed by Austin Model-1 with 10000 as maximum number of cycles, 0.01 as convergence criteria (root mean square gradient) and 1.0 as constant (medium’s dielectric constant which is 1 for in vacuo) in dielectric properties. The default values of 30.0 and 10.0 Kcal/mol were used for electrostatic and steric energy cutoff.
Chemical structures of 2-substituted-N-[4-(1-methyl-4,5- diphenyl-1H-imidazole-2-yl)phenyl] acetamide derivatives
PPT Slide
Lager Image
Chemical structures of 2-substituted-N-[4-(1-methyl-4,5- diphenyl-1H-imidazole-2-yl)phenyl] acetamide derivatives
IC50represent inhibition activities of compounds 1−18 against breast MCF-7 cell line respectively
PPT Slide
Lager Image
IC50 represent inhibition activities of compounds 1−18 against breast MCF-7 cell line respectively
- 2D-QSAR Analysis
Calculation of descriptors
Number of descriptors was calculated after optimization or minimization of the energy of the data set molecules. Various types of physicochemical descriptors were calculated: Individual (Molecular weight, H-Acceptor count, H-Donor count, XlogP, slogP, SMR, polarisablity, etc.), retention index (Chi), atomic valence connectivity index (ChiV), Path count, Chi chain, ChiV chain, Chain Path- Count, Cluster, Pathcluster, Kappa, Element count (H, N, C, S count etc.), Distance based topological (DistTopo, ConnectivityIndex, WienerIndex, BalabanIndex), Estate numbers (SsCH3count, SdCH2count, SssCH2count, StCHcount, etc.), Estate contribution (SsCH3-index, SdCH2- index, SssCH2-index, StCH index), Information theory based (Ipc, Id etc.) and Polar surface area. More than 200 alignment independent descriptors were also calculated using the following attributes. A few examples are T_2_O_7, T_N_N_5, T_2_2_6, T_C_O_1, T_O_Cl_5 etc. The invariable descriptors (the descriptors that are constant for all the molecules) were removed, as they do not contribute to QSAR.
Generation of training and test sets
In order to evaluate the QSAR model, data set was divided into training and test set using sphere exclusion, random selection and manual selection method. Training set is used to develop the QSAR model for which biological activity data are known. Test set is used to challenge the QSAR model developed based on the training set to assess the predictive power of the model which is not included in model generation.
Sphere Exclusion Method: In this method initially data set were divided into training and test set using sphere exclusion method. In this method dissimilarity value provides an idea to handle training and test set size. It needs to be adjusted by trial and error until a desired division of training and test set is achieved. Increase in dissimilarity value results in increase in number of molecules in the test set.
Random Selection Method: In order to construct and validate the QSAR models, both internally and externally, the data sets were divided into training [90−60% (90%, 85%, 80%, 75%, 70%, 65% and 60%) of total data set] and test sets [10−40% (10%, 15%, 20%, 30%, 35% and 40%) of total data set] in a random manner. 10 trials were run in each case.
Manual Data Selection Method: Data set is divided manually into training and test sets on the basis of the result obtained in sphere exclusion method and random selection method.
Generation of 2D-QSAR models
PLSR was used for model generation. PLSR is an expansion of the multiple linear regression (MLR) models. In its simplest form, a linear model specifies the (linear) relationship between a dependent (response) variable and a set of predictor variables. PLSR extends MLR without imposing the restrictions employed by discriminant analysis, principal component regression (PCR) and canonical correlation. In PLSR, prediction functions are represented by factors extracted from the Y’XX’Y matrix. The number of such prediction functions that can be extracted typically will exceed the maximum of the number of Y and X variables. PLSR is probably the least restrictive of the various multivariate extensions of the multiple linear regression models. This flexibility allows it to be used in situations where the use of traditional multivariate methods is severely limited, such as when there are fewer observations than predictor variables. PLSR can be used as an exploratory analysis tool to select suitable predictor variables and to identify outliers before classical linear regression. All the calculated descriptors were considered as independent variable and biological activity as dependent variable.
- 3D-QSAR Analysis
kNN-MFA
kNN-MFA is novel methodology, unlike conventional QSAR regression methods; this methodology can handle nonlinear relationships of molecular field descriptors with biological activity, thus making it a more accurate predictor of biological activity. Conventional correlation methods try to generate linear relationship with the activity, where as kNN is inherently non-linear method and is better able to explain activity trends. The kNN technique is a conceptually simple approach to pattern recognition problems. In this method, an unknown pattern is classified according to the majority of the class memberships of its k nearest neighbors in the training set. The nearness is measured by an appropriate distance metric (e.g., a molecular similarity measure, calculated using field interactions of molecular structures). The standard kNN method is implemented simply as follows: (i) calculate distances between an unknown object (u) and all the objects in the training set; (ii) select k objects from the training set most similar to object u, according to the calculated distances; (iii) classify object u with the group to which a majority of the k objects belong. An optimal k value is selected by the optimization through the classification of a test set of samples or by the leaveone out cross-validation. The variables and optimal k values are chosen using different variable selection methods as described below.
kNN-MFA with Simulated Annealing:
Simulated Annealing (SA) is another stochastic method for function optimization employed in QSAR. Simulated annealing (SA) is the simulation of a physical process, ‘annealing’, which involves heating the system to a high temperature and then gradually cooling it down to a preset temperature (e.g., room temperature). During this process, the system samples possible configurations distributed according to the Boltzmann distribution so that at equilibrium, low energy states are the most populated.
kNN-MFA with Stepwise (SW) Variable Selection:
This method employs a stepwise variable selection procedure combined with kNN to optimize the number of nearest neighbors (k) and the selection of variables from the original pool as described in simulated annealing.
kNN-MFA with Genetic Algorithm:
Genetic algorithms (GA) first described by Holland mimic natural evolution by modeling a dynamic population of solutions. The members of the population, referred to as chromosome, encode the selected features. The encoding usually takes form of bit strings with bits corresponding to selected features set and others cleared. Each chromosome leads to a model built using the encoded features. By using the training data, the error of the model is quantified and serves as a fitness function. During the course of evolution, the chromosomes are subjected to crossover and mutation. By allowing survival and reproduction of the fittest chromosomes, the algorithm effectively minimizes the error function in subsequent generations.
Alignment rules
Molecular alignment was used to visualize the structural diversity in the given set of molecules. This was followed by generation of common rectangular grid around the molecules. The template structure was used for alignment by considering the common elements of the series as shown in . 1 . The reference molecule 28 is chosen high inhibitory effect which made it a valid lead molecule and therefore was chosen as a reference molecule. After optimizing, the template structure and the reference molecule were used to superimpose all molecules from the series using the template alignment method. 29 kNN-MFA method requires suitable alignment of given set of molecules after optimization; alignment was carried out by template based alignment method. Stereo view of aligned molecules in training set and test set is shown in . 2 .
PPT Slide
Lager Image
2-Substituted-N-[4-(1-methyl-4,5-diphenyl-1H-imidazole- 2-yl)phenyl] acetamide derivatives (template structure).
PPT Slide
Lager Image
3D view of aligned 2-substituted-N-[4-(1-methyl-4,5- diphenyl-1H-imidazole-2-yl)phenyl] acetamide derivatives.
Creation of interaction energies
Methyl probe with charge 1 and energy cut-off for electrostatic 10 Kcal/mol and for steric 30 Kcal/mol, dielectric constant 1 and charge type Gasteiger-marsili were used to calculate steric and electrostatic fields Gasteiger-marsili, et al. 28 The fields were computed at each lattice intersection of a regularly spaced grid of 2.0Å within defined three-dimensional region.
Generation of training and test sets
In order to evaluate the QSAR model, data set was divided into training and test set using sphere exclusion, random selection and Manual selection method. Training set is used to develop the QSAR model for which biological activity data are known. Test set is used to challenge the QSAR model developed based on the training set to assess the predictive power of the model which is not included in model generation.
RESULTS AND DISCUSSION
- 2D-QSAR Models
Different sets of 2D-QSAR models were generated using the Multiple Linear Regression analysis in conjunction with stepwise forward-backward variable selection method. Different training and test set were constructed using random and manual selection method. Training and test set were selected if they follow the unicolumn statistics, i.e., maximum of the test is less than maximum of training set and minimum of the test set is greater than of training set, which is prerequisite for further QSAR analysis. This result shows that the test is interpolative i.e., derived from the min-max range of training set. The mean and standard deviation of the training and test set provides insight to the relative relative difference of mean and point density distribution of the two sets. The statistical significant 2D-QSAR models are given in 3 .
The selection of the best model is based on the values of r 2 (squared correlation coefficient), q 2 (cross-validated correlation coefficient), pred_r 2 (predicted correlation coefficient for the external test set), F (Fisher ratio) reflects the ratio of the variance explained by the model and the variance due to the error in the regression. High values of the F-test indicate that the model is statistically significant. r 2 se, q 2 se and pred_r 2 se are the standard errors terms for r 2 , q 2 and pred_r 2 respectively. The statistically significant 2D-QSAR model is shown as follows.
Model-1 (Test set: 10, 13, 18, 7, 8, 9) pIC50 = −0.2392 (T_2_N_1) + 0.2570 (SsCH3count) + 2.8148
Statistics: [n = 12; Degree of freedom = 9; r2 = 0.8901; q2 = 0.8130; F test = 36.4635; r2se = 0.1696;q2se = 0.2212; pred_r2 = 0.4229; pred_r2se = 0.4606]
Model-2 (Test size: 1, 13, 14, 15, 6, 8) pIC50=−0.4102(T_N_S_3)− 0.5910 (T_N_CL_6)− 0.1347 (SaaaC count) + 1.9130
Statistics: [n = 12; Degree of freedom = 8; r2 = 0.9462; q2 = 0.6915, F test = 46.8763; r2se = 0.1084; q2se = 0.2595; pred_r2 = 0.3655; pred_r2se = 0.5573]
In the above QSAR equations, n is the number of molecules (Training set) used to derive the QSAR model, r 2 is the squared correlation coefficient, q 2 is the cross-validated correlation coefficient, pred_r 2 is the predicted correlation coefficient for the external test set, F is the Fisher ratio, reflects the ratio of the variance explained by the model and the variance due to the error in the regression. High values of the F-test indicate that the model is statistically significant. r 2 se, q 2 se and pred_r 2 se are the standard errors terms for r 2 , q 2 and pred_r 2 (smaller is better).
Statistical evaluation of 2D-QSAR models
PPT Slide
Lager Image
Statistical evaluation of 2D-QSAR models
- Interpretation of the Models
Model-1
From equation, model-1 explains 89.01% (r 2 = 0.8901) of the total variance in the training set as well as it has internal (q 2 ) and external (pred_r 2 ) predictive ability of 81.30% and 42.29% respectively. The F-test shows the statistical significance of 99.99% of the model which means that probability of failure of the model is 1 in 10000. In addition, the randomization test shows confidence of 99.9999 (Alpha Rand Pred R^2 = 0.00000) that the generated model is not random and hence chosen as the QSAR model. From QSAR model-1, negative coefficient value of T_2_N_1 [count of number of double bonded atoms (i.e., any double bonded atom, T_2) separated from nitrogen atom by single bonds] on the biological activity indicated that lower values leads to good inhibitory activity while higher value leads to reduced inhibitory activity while positive coefficient value of SsCH 3 count[the total number of CH 3 group connected with single bond] on the inhibitory activity indicated that higher value leads to better inhibitory activity whereas lower value leads to decrease inhibitory activity.
Contribution chart for model-1 is represented in . 3 reveals that the descriptors SsCH 3 count, contributing 25.70% respectively. Another descriptors T_2_N_1 is contributing inversely 23.92% respectively to biological activity.
PPT Slide
Lager Image
Plot of contribution chart for 2D-QSAR model-1.
Data fitness plot for model-1 is shown in . 4 . The plot of observed vs predicted activity provides an idea about how well the model was trained and how well it predicts the activity of external test set.
PPT Slide
Lager Image
Data fitness plot for 2D-QSAR model-1.
The graph of observed vs predicted activity of training and test sets for model-1 is shown in . 5(a) and 5(b) , it reveals that the model is able to predict the activity of training set quite well as well as external test set, providing Confidence of model. Result of the observed and predicted inhibitory activity for the training and test compounds for the Model-1 is shown in 5 .
PPT Slide
Lager Image
(a) Plot of observed versus predicted activity of 2D-QSAR model-1 (training set), (b) Plot of observed versus predicted activity of 2D-QSAR model-1 (test set).
From equation, Model-2 explains 94.62% (r 2 = 0.9462) of the total variance in the training set as well as it has internal (q 2 ) and external (pred_r 2 ) predictive ability of 69.15% and 36.55% respectively. From QSAR model-2, it was observed that Negative coefficient value of T_2_N_0 [count of number of double bonded atoms (i.e., any double bonded atom, T_2) separated from carbon atom], T_N_CL_6 [count of number of nitrogen atoms (single, double or triple bonded) separated from any other chlorine atom (single or double or triple bonded) by 6 bonds in a molecule and SaaaCcount (total number of carbon connected with three aromatic bonds) on the biological activity indicated that lower values leads to good inhibitory activity while higher value leads to reduced inhibitory activity.
Statistical evaluation of 3D-QSAR models
PPT Slide
Lager Image
Statistical evaluation of 3D-QSAR models
PPT Slide
Lager Image
Plot of contribution chart for 2D QSAR model-2.
Contribution chart for model-2 is represented in . 6 , it reveals that the descriptors T_2_N_0, T_N_CL_6, SaaaCcount contributing 41.02%, 59.10% and 13.47% respectively.
PPT Slide
Lager Image
Data fitness plot for 2D QSAR model-2.
Data fitness plot for model-2 is shown in . 7 ; the plot of observed vs predicted activity provides an idea about how well the model was trained and how well it predicts the activity of external test set.
The graph of observed vs predicted activity of training and test sets for model-2 is shown in . 8(a) and 8(b) , reveals that the model is able to predict the activity of training set quite well as well as external test set, providing confidence of model. Result of the observed and predicted inhibitory activity for the training and test compounds for the Model-2 is shown in 5 .
PPT Slide
Lager Image
(a) Plot of observed versus predicted activity of 2D-QSAR model-2 (training set), (b) Plot of observed versus predicted activity of 2D-QSAR model-2 (test set).
Actual and predicted activities for 18 compounds based on the best 2D/3D-QSAR models
PPT Slide
Lager Image
Actual and predicted activities for 18 compounds based on the best 2D/3D-QSAR models
- 3D-QSAR Model
kNN-MFA samples the steric and electrostatic fields surrounding a set of ligands and constructs 3D-QSAR models by correlating these 3D fields with the corresponding biological activities. The statistical significant 3D-QSAR models 4 .
The selection of the best model is based on the values of q 2 (internal predictive ability of the model) and that of pred_r 2 (the ability of the model to predict the activity of external test set). The statistical significant 3D-QSAR models for p IC 50 (model-1) and p IC 50 (model-2) are given below.
Model-1(3D)
p IC 50 = −S_2012(−0.0792, −0.0467) − S_2567(−0.0566, −0.0152) − S_1620(−0.0789, −0.0441) + E_3168(0.0857, 0.0773)
Statistics:
[kNN = 2; n = 12; Degree of freedom = 7; q 2 = 0.9372; q 2 _se = 0.1304; pred_r 2 = 0.6574; pred_r 2 se = 0.2777]
The model 1 explains values of k (2), q 2 (0.9372), pred_r 2 (0.6574), q 2 _se (0.1304), and pred_r 2 se (0.2777) prove that QSAR equation so obtained is statistically significant and shows the predictive power of the model is 93.72% (internal validation). 5 represents the predicted inhibitory activity by the model-1 for training and test set.
PPT Slide
Lager Image
Data fitness plot for 3D-QSAR model-1.
The data fitness plot for model-1 is shown in . 9 . The plot of observed vs. predicted activity provides an idea about how well the model was trained and how well it predicts the activity of the external test set.
From . 10(a) and 10(b) it can be seen that the model is able to predict the activity of the training set quiet well as well as external test set, providing confidence of the model.
Result plot in which 3D-alignment of molecules with the important steric and electrostatic points contributing in the model-3 with ranges of values shown in the parenthesis represented in . 11 . It shows the relative position and ranges of the corresponding important steric and electrostatic fields in the model provides guideline for new molecule design as follows.
  • (a) Electrostatic field, E_3168 (0.0857, 0.0773) has positive range indicates that positive electrostatic potential is favorable for increase in the activity and hence less electronegative substituent group is preferred in that region.
  • (b) Steric filed, S_2012 (−0.0792, −0.0467) has negative range indicates that negative steric potential is favorable for increase in the activity and hence less bulky substituent group is preferred in that region.
  • (c) Steric filed, S_2567 (−0.0566, −0.0152) also has negative range indicates that negative steric potential is favorable for increase in the activity and hence less bulky substituent group is preferred in that region.
  • (d) Steric filed, S_1620 (−0.0789, −0.0773) also has negative range indicates that negative steric potential is favorable for increase in the activity and hence less bulky substituent group is preferred in that region.
PPT Slide
Lager Image
a) Plot of observed versus predicted activity of 3D-QSAR model-1 (training set), (b) Plot of observed versus predicted activity of 3D-QSAR model-1 (test set).
PPT Slide
Lager Image
Plot of contribution chart for 3D QSAR model-1.
Model-2(3D)
p IC 50 =−E_1712(−0.9546, −0.7371)+S_1633(0.1019, 0.1084)
Statistics:
[kNN = 2; n = 12; Degree of freedom = 9; q 2 = 0.8561; q 2 _se = 0.1863; pred_r 2 = 0.5311; pred_r 2 se = 0.3699]
PPT Slide
Lager Image
Data fitness plot for 3D QSAR model-2.
In model 2(3D), values of kNN (2), q 2 (0.8561), pred_r 2 (0.5311), q 2 se (0.1863), and pred_r 2 se (0.3699) prove that QSAR equation so obtained is statistically significant and shows the predictive power of the model is 85.61% (internal validation). 5 represents the predicted inhibitory activity by the model-2 (3D) for training and test set.
The data fitness plot for model-2(3D) is shown in . 12 . The plot of observed vs predicted activity provides an idea about how well the model was trained and how well it predicts the activity of the external test set.
From . 13(a) and 13(b) it can be seen that the model is able to predict the activity of the training set quiet well as well as external test set, providing confidence of the model.
Result plot in which 3D-alignment of molecules with the important steric and electrostatic points contributing in the model with ranges of values shown in the parenthesis represented in . 14 . It shows the relative position and ranges of the corresponding important steric and electrostatic fields in the model provides guideline for new molecule design as follows.
  • (a) Electrostatic field, E_1712 (−0.9546, −0.7371) has negative range indicates that negative electrostatic potential is favorable for increase in the activity and hence more electronegative substituent group is preferred in that region.
  • (b) Steric field, S_1633 (0.1009, 0.1084) has positive range indicates that positive steric potential is favorable for increase in the activity and hence more bulky substituent group is preferred in that region.
PPT Slide
Lager Image
a) Plot of observed versus predicted activity for 3D-QSAR model-2 (training), (b) Plot of observed versus predicted activity of 3D QSAR model-2 (test set).
PPT Slide
Lager Image
Plot of contribution chart for 3D-QSAR model-2.
CONCLUSION
Statistically significant 2D/3D-QSAR models were generated with the purpose of deriving structural requirements for the inhibitory activities of some imidazole-(benz)azole derivatives against breast MCF-7 cell lines. The validation of 2D-QSAR models was done by the cross-validation test, randomization tests and external test set prediction. The best 2D-QSAR models indicate that the descriptors of SsCH3count, T_2_N_1 influenced the inhibition activity. kNN-MFA investigated the substitutional requirements for the receptor-drug interaction and constructed the best 3DQSAR models by Multiple Linear Regression method, providing useful information in characterization and differentiation of their binding sites. In conclusion, the information provided by the robust 2D/3D-QSAR models use for the design of new molecules ( 6 ) and hence, this method is expected to provide a good alternative for the drug design.
Structure and predicted activity of newly designed compounds
PPT Slide
Lager Image
*Increased predicted activity.
Acknowledgements
The authors are thankful to V-Life Sciences Technologies Private Limited, 1 Akshay, 50 Anand Park, Aundh, Pune, India for providing trial version of the software. And the publication cost of this paper was supported by the Korean Chemical Society.
References
Krikas G. , Schulpis K. , Reclos G. , Kokotos G. 1998 Clin. Biochem. 31 405 -    DOI : 10.1016/S0009-9120(98)00016-2
Waring M. , Bailly C. 1994 Gene 149 69 -    DOI : 10.1016/0378-1119(94)90414-6
Rehn C. , Pindur U. 1996 Monatsh. Chem. 127 631 -    DOI : 10.1007/BF00817255
Baginski M. , Fogolari F. , Briggs J. 1997 J. Mol. Biol. 274 253 -    DOI : 10.1006/jmbi.1997.1399
Shui X. 2000 Curr. Med. Chem. 7 59 -    DOI : 10.2174/0929867003375470
Neidle S. 2008 In Cancer Drug Desing and Discovery 2nd Ed Elsevier Inc U.S.A
Moore M. , Hunter W. , Langlis-D'Estaintot B. , Kennard O. 1989 J. Mol. Bio. 206 693 -    DOI : 10.1016/0022-2836(89)90577-9
Pindur U. , Haber M. , Sattler K. 1993 J. Chem. Educ. 70 263 -    DOI : 10.1021/ed070p263
Zhigang L. , Qing Y. , Xuhong Q. 2005 Bioorg. Med. Chem. Lett. 15 3143 -    DOI : 10.1016/j.bmcl.2005.04.012
Claiborne C. , Liverton N. , Nguyen K. 1998 Tetrahedron Lett. 39 8939 -    DOI : 10.1016/S0040-4039(98)02058-9
Cui B. , Zheng B. , He K. , Zheng Q. 2003 J. Nat. Prod. 66 1101 -    DOI : 10.1021/np030031i
Fazeny-Dorner B. , Veitl M. , Wenzel C. , Piribauer M. , Rossler K. , Dieckmann K. , Ungersbock K. , Marosi C. 2003 Anticancer Drugs. 14 437 -    DOI : 10.1097/00001813-200307000-00008
Haluska P. , Dy G. , A. Adjei A. 2008 Eur. J. cancer. 38 1685 -
Kotteas E. , Alamara C. , Kiagia M. , Pantazopoulos K. , Oufas A. , Provata A. , Chapidou A. , Syrigos K. N. 2008 Anticancer Res. 28 529 -
Perez-ruixo J. , Piotrovskij V. , Zhang S. , Hayes S. , DePorre P. , Zannikos P. 2006 Br. J. Clin, Pharmacol. 62 81 -    DOI : 10.1111/j.1365-2125.2006.02615.x
Krezel I. 1998 Farmaco 53 342 -    DOI : 10.1016/S0014-827X(98)00031-7
Kumar D. , Jacob M. R. , Reynolds M. B. , Kerwin S. M. 2002 Bioorg. Med. Chem. 10 3997 -    DOI : 10.1016/S0968-0896(02)00327-9
Hose C. , Hollingshead M. , Sausville E. , Monks A. 2003 Mol. Cancer Ther. 2 1265 -
Al-Masoudi A. , Al-Soud Y. , Al-Salihi N. , Al-Masoudi N. 2006 Chem. Hetetrocycl. Compd. 42 1377 -    DOI : 10.1007/s10593-006-0255-3
Jackman A. , Kimbell R. , Aherne G. , Brunton L. , Jansen G. , Stephens T. , Smith M. , Wardleworth J. , Boyle F. 1997 Clin. Cancer Res. 3 911 -
Nelson J. , Rose L. , Benneft L. 1977 Cancer Res. 37 182 -
Martin B. , Mann J. , Sageot A. 1999 J. Chem. Soc. Perkin Trans. 1 2455 -
Brian B. , Andrew M. , Hana K. , Padmakumari T. , William P. , Jack C. 1997 Mol. Pharmacol. 52 839 -    DOI : 10.1124/mol.52.5.839
Hansch C. , Kurup A. , Garg R. , Gao H. 2001 Chem. Rev. 101 619 -    DOI : 10.1021/cr0000067
Yusuf O. , Ilhan I. , Zerrin I. , Gulsen A. 2010 Eur. J. Med. Chem. 45 3320 -    DOI : 10.1016/j.ejmech.2010.04.015
Golbraikh A. , Tropsha A. 2002 J. Comput. Aided Mol. Des. 16 357 -    DOI : 10.1023/A:1020869118689
2012 VLifeMDS 4.2; Molecular Design Suite Vlife Sciences Technologies Pvt. Ltd Pune, India
Gasteiger J , Marsili M. 1980 Tetrahedron. 36 3219 -    DOI : 10.1016/0040-4020(80)80168-2
Ajmani S. , Jhadav K. , Kulkarni S. A. 2006 J. Chem. Inf. Model. 46 24 -    DOI : 10.1021/ci0501286