RHAZES: Green and Applied Chemistry

Vol.3, 12, 2018, pp. 33-46

ISSN: 2605-6895

In silico analysis of 3D-QSAR and molecular docking results to

design new 1,2,4-triazoles as antibacterial agents
Adib GHALEB1*, Adnane AOUIDATE1, Mounir GHAMALI1, Samir CHTITA1, Abdellah OUSAA1,
Mohammed BOUACHRINE2, Tahar LAKHLIFI1, Abdelouahid SBAI1
1
MCNS, Faculty of Science, Moulay Ismail University, Meknes, Morocco
1, 2
MEM, EST, Moulay Ismail University, Meknes, Morocco

Article Info ABSTRACT

Drug screening is a long and costly process confronted with low
Article history:
productivity and challenges in using animals, which limit the
Received 12/08/2018 discovery of new drugs, for this reason an efforts have been
Accepted 30/10/2018 dedicated to develop3D-QSAR (comparative molecular field analysis
(CoMFA) and comparative molecular similarity indices analysis
(CoMSIA)) model and Surflex-docking study, to design new
compounds. In this work CoMFA and CoMSIA models using 22
Keyword: compounds in the training set gave Q2 values of 0.83 and 0.86, and
3D-QSAR r2 values of 0.91 and 0.82, respectively. The adapted alignment
CoMFA method with the suitable parameters resulted in reliable models.
CoMSIA Contour maps produced by the CoMFA and CoMSIA models were
Surflex-docking employed to rationalize the key structural requirements responsible
Anti-bacterial for the activity, while Surflex-docking revealed the interactions of
1,2,4triazole triazole compounds with receptor; some essential features were also
identified. Based on the 3D-QSAR and docking results, a set of new
molecules with high predicted activities were designed. The total
scoring of inactive, active and proposed compounds were compared
to each other to determine the stability of ligand in receptor.

Corresponding Author:
Adress Email: adib.ghaleb@gmail.com
Phone: +212658626168

1. INTRODUCTION

Triazole derivatives possess a large significance in medicinal chemistry and numerous

heterocyclic triazole compounds having different biological activities. Triazole is a five
membered basic compound, containing two carbon and three nitrogen atoms having
molecular formula C2H3N3. Triazole derivatives possess wide variety of pharmacological
activities such as antifungal, antibacterial, antiviral, anticancer, anticonvulsant, anti-
inflammatory, antioxidant, anti-tubercular, anti-malarial, anti-nociceptive and other

33
anticipated activities [1-2]. Various substituted triazole compounds are responsible for
producing antibacterial effects and further modifications can be made on it to enhance their
antimicrobial effects[3-4]. The studied compounds were screened for their antibacterial
activity against Escherichia coli bacterial stains by serial plate dilution method [5]. Serial
dilutions of the drug in Muller-Hinton broth were taken in tubes and their pH was adjusted to
5.0 using phosphate buffer. A standardized suspension of the test bacterium was inoculated
and incubated for 16-18 h at 37ºC. The minimum inhibitory concentration (MIC) was noted
by seeing the lowest concentration of the drug at which there was no visible growth [6].
Quantitative structure activity relationship (QSAR) is one of the major tools in drug discovery
to explore ligand– receptor/enzyme interactions; it correlate the specific biological activities
or physical properties of a series of compounds with the measured or computed molecular
properties of the compounds, in terms of descriptors. The development of a quantitative SAR
with the aid of various physicochemical parameters has been an important task in lead
optimization [7]. The present study aimed to explore 3D-QSAR and molecular docking
approaches to propose new triazoles compounds as antibacterial agents.

2. MATERIAL AND METHODS

2.1. Data collection:

For docking and 3D-QSAR analysis a set of 28compounds with reported MIC values as
antibacterial activity against Escherichia coli, were taken from literature [8]. For the QSAR
analysis the in vitro biological activities MIC were converted into the corresponding pMIC
values and are listed with their corresponding structures in Fig. 1 and Table 1, the dataset was
split into two sets, 22 molecules were chosen randomly to develop the quantitative model
(training set) and the rest were used to test the performance of the proposed model (test set).

1-19 20-28
Fig. 1: Chemical structures of the studied compounds

34
Table 1: Characterization data and antibacterial activities of Triazole derivatives.
No Ar R Mol. formula PMIC

1 C6H5 SCH3 C17H14N4OS2 2.20

2 2-ClC6H4 SCH3 C17H13ClN4OS2 1.60

3 4-CH3C6H4 SCH3 C18H16N4OS2 2.20

4 4-OCH3C6H4 SCH3 C18H16N4O2S2 2.20

5 4-NH2C6H4 SCH3 C17H15N5OS2 2.20

*6 2,3-(Cl)2C6H4 SCH3 C17H12Cl2N4OS2 2.20

7 C6H5CH2 SCH3 C18H16N4O2S2 2.20

*8 C6H5 SC2H5 C18H16N4OS2 2.20

9 2-ClC6H4 SC2H5 C18H15ClN4OS2 1.60

10 4-CH3C6H4 SC2H5 C19H18N4OS2 2.20

*11 4-OCH3C6H4 SC2H5 C19H18N4O2S2 1.60

*12 2-NH2C6H4 SC2H5 C18H17N5OS2 1.60

13 2,3-(Cl)2C6H4 SC2H5 C18H14Cl2N4OS2 1.60

14 2-OHC6H4 SC2H5 C18H16N4O2S2 1.60

15 C6H5 SO2CH3 C17H14N4O3S2 1.90

16 2-ClC6H4 SO2CH3 C17H13ClN4O3S2 1.90

17 3-ClC6H4 SO2CH3 C17H13ClN4O3S2 1.90

*18 CH2C6H5 SO2CH3 C18H16N4O3S2 1.90

19 2,3-Dichloro-C6H4 SO2CH3 C17H12Cl2N4O3S2 1.60

20 2-OH-benzamide SCH3 C19H17N5O3S2 2.20

*21 2,4-Dichloro-C6H3 SCH3 C18H14Cl2N4OS2 2.20

22 4-ClC6H4 SCH3 C18H15Cl2N4OS2 2.20

23 2-OH-benzamide SC2H5 C20H19N5O3S2 2.20

24 2,4-Dichloro-C6H3 SC2H5 C19H16Cl2N4OS2 1.60

25 4-ClC6H4 SC2H5 C19H17ClN4OS2 1.60

26 2-OH-benzamide SO2CH3 C19H17N5O5S2 1.90

27 2,4-Dichloro-C6H3 SO2CH3 C18H14Cl2N4O3S2 1.90

28 4-ClC6H4 SO2CH3 C18H15ClN4O3S2 1.60

35
 2.2. Minimization and alignment:

All structures were sketched with sketch module in SYBYL and optimized using Tripos force
field [9] with the Gasteiger–Huckel charges [10] by conjugated gradient methodwith gradient
convergence criteria of 0.01 kcal/mol. Simulated annealing on the optimized structures was
performed with 20 cycles. Molecular alignment is one of the most sensitive parameters in 3D-
QSAR analyses. In this work, all molecules were aligned on the common core, using the
simple alignment method in Sybyl [11].The most active compound 1 was used as template.
The superimposed structures of aligned data set are shown in Figure 2.

Aligned compound CORE

Fig. 2 : 3D-QSAR structure superposition and alignment of training set using molecule 1 as a
template.
2.3. 3D QSAR Studies

To understand and explore the contributions of electrostatic, steric and hydrophobic fields in
different compounds of the data set and to create predictive 3D QSAR models, CoMFA and
CoMSIA studies were performed based on the molecular alignment strategy, as previously
described in the literature [12].

2.3. CoMFA and CoMSIA

Based on the molecular alignment, CoMFA and CoMSIA studies were performed to analyze
the specific contributions of steric, electrostatic, and hydrophobic effects. The CoMFA
method allows calculating steric and electrostatic properties according to the Lennard-Jones
and Coulomb potentials, respectively. Using the CoMSIA approach, we calculated the
similarity indices in the space surrounding each of the molecules in the dataset. CoMFAsteric
and electrostatic interaction fields were calculated at each lattice intersection point of a
regularly spaced grid of 2.0 A˚. The default value of 30 kcal/mol was set as a maximum steric
and electrostatic energy cutoff [13]. With standard options for scaling of variables, the
regression analysis was carried out using the full cross-validated partial least squares (PLS)
method (leave-one-out) [14]. The minimum sigma (column filtering) was set to 2.0 kcal/molto
improve the signal-to-noise ratio by omitting the lattice points whose energy variation was
below this threshold. The final non-cross-validated model was developed using optimal
number of components that had both the highest Q2 value and the smallest value of standard
error predictions. The predictive r2 was used to evaluate the predictive power of the CoMFA
model and was based only on the test set. Several CoMFA models were built by considering
permutations of molecules between training and test sets. The best model amongst them was
chosen, based on high Q2, r2 values and small Standard Error of Estimate (SEE) value. In
CoMSIA, a distance-dependent Gaussian-type physicochemical property has been adopted to
36
avoid singularities at the atomic positions and dramatic changes of potential energy for grids
being in the proximity of the surface. With the standard parameters and no arbitrary cutoff
limits, five fields associated to five physicochemical properties, namely steric (S),
electrostatic (E), and hydrophobic effects (H) and hydrogen bond donor (D) and acceptor (A)
indices were calculated.

2.4. PLS analysis :

Partial least squares statistical method used in deriving the 3D-QSAR models is an extension
of multiple regression analysis in which the original variables are replaced by a small set of
their linear combinations. PLS method with leave-one-out (LOO) cross-validation was used in
this study to determine the optimal numbers of components using cross-validated coefficient
Q2. The external validation of various models was performed using a test set of five
molecules. The final analysis (non-cross-validated analysis) was carried out using the
optimum number of components obtained from the cross-validation analysis to get correlation
coefficient (R2). The Q2 value determines the internal predictive ability of the model while R2
value evaluates the internal consistency of the model. Thus, the best QSAR model was chosen
based on a combination of Q2 and R2.

2.5. Y-Randomization Test:

The obtained models were further validated by the Y-Randomization method. The Y vector (-
LogMIC) was randomly shuffled many times and after every iteration, which led to a new
QSAR model. The new QSAR models are expected to have lower Q2andR2 values than those
the original models. This technique was carried out to eliminate the possibility of the chance
correlation. If higher values of the Q2andR2 are obtained, it means that an acceptable 3D-
QSAR model cannot be generated for this dataset because of structural redundancy and
chance correlation.
2.6. Molecular Docking:

The Surflex-Dock [15] was applied to study molecular docking by using an empirical scoring
function and a patented search engine to dock ligands into a protein’s binding site. The
antimicrobial protein was retrieved from the RCSB Protein Data Bank (PDB entry code:
4G5G). The ligands were docked into corresponding protein’s binding site using an empirical
scoring function and a patented search engine in Surflex-Dock [15]. All water molecules in
the protein were deleted and the polar hydrogen atoms added. Protomol, which is a
computational representation of ligand making every potential interaction with the binding
site, is applied to guide molecular docking and predicted binding mode. All protomols could
be established by three ways: (1) automatic: Surflex-Dock finds the largest cavity in the
receptor protein; (2) ligand-based: a ligand in the same coordinate space as the receptor; (3)
residues based: specified residues in the receptor [16,17].
In this study, we applied automatic docking: the 4G5Gstructure was utilized in the subsequent
docking experiments without energy minimization with other parameters established by
default in the software. Surflex-Dock scores (total scores) were expressed in -log10(Kd) units
to represent binding affinities. Then, the MOLCAD (Molecular Computer Aided Design)
program was employed to visualize the binding mode between the protein and ligand.
MOLCAD calculates and exhibits the surfaces of channels and cavities, as well as the
separating surface between protein subunits [18-20]. MOLCAD program provides several
37
capabilities to create a molecular surface [15]. The fast Connolly method using a marching
cube algorithm to engender the surface was applied in this work, thus the MOLCAD Robbin
and Multi-Channel surface program exhibited with copious potentials was used. Moreover,
Surflex-Dock total scores, which represent binding affinities, were applied to estimate the
ligand-receptor interactions of newly designed molecules. Every optimized conformation of
each molecule in the data set was energetically minimized employing the Tripos force field
and the Powell conjugated gradient algorithm with a convergence criterion of 0.05 kcal/mol
A˚ and Gasteiger-Huckelcharges [10].

3. RESULTS AND DISCUSSION

3.1. CoMFA results:

Results of Table 2 demonstrate that CoMFA model has high R2 (0.91) and F (19.12) values,
and a small Scv(0.12) as well as cross-validated correlation coefficient Q2 (0.83) with two as
optimum number of components. The external predictive capability of a QSAR model is
generally cross checked and validated using test sets. The five randomly selected test setswere
optimized and aligned in the same manner as training sets. The external validation gave high
value of r2ext (0.92) which indicates that prediction ability of CoMFA model is acceptable. The
rations of steric to electrostatic contributions were found to be 61:49, which indicates
thatsteric interactions are much more important than electrostatic.

3.2.CoMSIA results :

3D-QSAR model was proposed based on CoMSIAdescriptor to explain and predict

quantitatively the Hydrophobic, Electrostatic, steric, donor and acceptor fields effects of
substituents on anti-bacterial activity of 28 compounds.
Different combinations of five fields were generated. The best CoMISA proposed model
contains four fields (Steric, Electrostatic, Hydrophobic, and Acceptor), the cross-validated
correlation coefficient Q2 value of the training set and non-cross-validated correlation
coefficient R2 are 0.86 and 0.82, respectively. The optimal number of principal components
using to generate the CoMSIA model is three, which is reasonable considering the number of
molecules used to build the model. The standard error was 0.16.
Finally, the prediction ability of the proposed model confirmed by using the external
validation, the r2ext value obtained is 0.89. These statistical results indicate the good stability
and the powerful predictive ability of CoMSIA model presented in Table 3.
Table 2: PLS Statistics of CoMFA and CoMSIA models.
Fractions
Model Q2 R2 Scv F N rext2
Ster Elec Acc Don Hyd

CoMFA 0.83 0.91 0.12 19.12 2 0,92 0.733 0.267 - - -

CoMSIA 0.86 0,82 0.16 16.13 3 0,89 0.204 0.426 0.072 0.00 0.298

Q2 : Cross-validated correlation coefficient.

N : Optimum number of components.
R2 : Non-cross-validated correlation coefficient.
Scv : Standard error of the estimate.
38
F : F -test value
rext2 : External validation correlation coefficient.

Table 3: Anti-bacterial and the predicted activities of Triazolederivatives.

CoMFA CoMSIA
No
PMIC predicted Residuals Predicted Residuals

1 2.20 2.24 0.04 2.23 0.03

2 1.60 1.55 0.05 1.58 0.02

3 2.20 2.16 0.04 2.23 -0.03

4 2.20 2.19 0.01 2.17 0.03

5 2.20 2.22 -0.02 2.17 0.03

*6 2.20 2.28 -0.08 2.11 0.09

7 2.20 2.19 0.01 2.22 -0.02

*8 2.20 2.26 -0.06 2.31 -0.11

9 1.60 1.63 -0.03 1.51 0.09

10 2.20 2.27 -0.07 2.28 -0.08

*11 1.60 1.64 -0.04 1.66 -0.06

*12 1.60 1.67 -0.07 1.47 0.13

13 1.60 1.63 -0.03 1.51 0.09

14 1.60 1.58 0.02 1.59 0.01

15 1.90 1.89 0.01 1.77 0.13

16 1.90 1.82 0.08 1.76 0.14

17 1.90 1.82 0.08 1.98 0.08

*18 1.90 1.81 -0.09 1.79 0.11

19 1.60 1.67 -0.07 1.63 0.03

20 2.20 2.23 -0.03 2.28 -0.08

*21 2.20 2.11 -0.09 2.08 0.12

22 2.20 2.22 -0.02 2.14 0.06

23 2.20 2.27 -0.07 2.19 0.01

24 1.60 1.62 -0.02 1.66 0.06

25 1.60 1.58 0.02 1.57 0.03

26 1.90 1.93 -0.03 1.88 0.02

27 1.90 1.81 0.09 1.96 -0.06

28 1.60 1.63 -0.03 1.59 0.01

39
* Test set molecules

3.3.Graphical Interpretation of CoMFA and CoMSIA results:

The CoMFA and CoMSIA contour maps were generated to rationalize the regions in 3D
space around the molecules where changes in each field were predicted to increase or
decrease the activity. The CoMFA steric and electrostatic contour maps are shown in Figure
3. Steric, electrostatic, hydrophobic and hydrogen bond acceptor contour maps of CoMSIA
are shown in Figure 4. Compound 1 was used as reference structure. All the contours
represented the default 80% and 20% level contributions for favored and disfavored regions,
respectively.

CoMFA Contour Maps

CoMFA electrostatic interactions represented by red and blue colored contours while steric
interactions represented by green and yellow colored contours. The bulky substituents are
favored in the green regions, and at yellow regions, they are unfavored. The bleu regions
indicate that positive charges are favored.

a) B)

Fig. 3 :Std* coeff. contour maps of CoMFA analysis with 2 Å grid spacing in combination
with compound 14. a) Steric fields: green contours (80% contribution) indicate regions where
bulky groups increase activity, while yellow contours (20% contribution) indicate regions
where bulky groups decrease activity.b). Electrostatic fields: blue contours (80%
contribution) indicate regions where groups with negative charges increase activity, while red
contours (20% contribution) indicate regions where groups with positive charges increase
activity.

Figure 3 shows a green contour around phenyl group, which indicate that bulky groups are
favored. Therefore, relying on CoMFA electrostatic contours, electron donor groups R group
can increase the activity, while around phenyl decrease the activity.

CoMSIA Contour Map:

The CoMSIA steric and electrostatic field contour maps indicate that bulky groups are not
favored in ortho and meta positions except for para position, also the absence of red contour
means that only electron donor groups are favored to increase the activity (Figure 4). This is
similar to CoMFA contour maps results.

40
The huge red contour around the phenyl group indicates that donor groups and hydrogen bond
donor groups could increase the activity, while the purple contour indicates that only groups
with hydrogen bond acceptor character can increase the activity. These results can explain
why compounds 2with small electro donating grouphave low activities, while compounds
with bulky electro withdrawing groups have high activities.

A)

B)

Fig. 4. Std* coeff. contour maps of CoMSIA analysis with 2 Å grid spacing for compound
14. a) Steric fields: green contours (80% contribution) indicate regions where bulky groups
increase activity, while yellow contours (20% contribution) indicate regions where bulky
groups decrease activity. b) Electrostatic fields: blue contours (80% contribution) indicate
regions where electron-donating groups increase activity, while red contours (20% C)
contribution) indicate regions where electron-withdrawing groups increase activity.c) H-bond
D)
acceptor fields: The purple (80% contribution) and red (20% contribution) contours favorable
and unfavorable positions for hydrogen bond acceptors respectively. d) Hydrophobic fields:
yellow contours (80% contribution) indicate regions where hydrophobic properties were
favored, while white contours (20% contribution) indicate regions hydrophilic properties were
favored.

3.4.Y-Randomization:

Table 4: Q2 and r2 values after several Y-randomization tests

CoMFA CoMSIA
Iteration
Q2 r2 Q2 r2
1 -0.31 0.57 0.21 0.65

2 -0.20 0.82 -0.17 0.72

3 0.11 0.60 0.18 0.46

4 0.14 0.52 -0.19 0.72

5 0.24 0.44 0.35 0.32

41
The Y-Randomization method was carried out to validate the CoMFA and CoMSIA models.
Several random shuffles of the dependent variable were performed; after every shuffle, a 3D-
QSAR was developed and the obtained results are shown in Table 4. The low Q2and r2 values
obtained after every shuffle indicate that the good result in our original CoMFA and CoMSIA
models are notA)due to a chance correlation of the training set.
3.2. New designed compounds

Based on CoMFA, CoMSIAmodelsresults; new molecules have been designed to enhance the
activity (Table 5). These compounds were minimized and aligned to the database using
compound 1 as a template. The best newly predicted structure X1 showed higher activity than
compound 1 (the most active compound of the series).
Surflex-docking is performed to confirm the predicted results of 3D-QSAR model.

Table 5: Chemical structure of newly designed molecules and their predicted pMIC based on
CoMFA and CoMSIA 3D-QSAR models.
Structure PredictedpIC50
No
Ar R CoMFA CoMSIA
X1 2,3,4(CH3)C6H4 SCH3 2.89 2.79
X2 4-tert-butylC6H4 SCH3 2.34 2.26

X3 2,3OH C6H4 SCH3 2.24 2.23

3.5. Docking results:

Molecular docking protocols are widely used to investigate the binding modes between the
ligand derivatives and receptor, which help understanding the 3D-QSAR study revealed by
CoMFA/CoMSIA models. The target ligand taken from the crystal structure was re-docked
into the active site to validate the accuracy of molecular docking. The figure5 shows the
different top 10 positions of molecule 1 in protein pocket and present the stable conformation
with scoring 3.06 higher than compound 2 (inactive molecule) with 1.31 scoring.

42
Fig. 5: The MOLCAD surface of the allosteric site within the compound 15.A) The stable
position of compound 15 in the receptor. B) The different ten positions of compound 15 in the
receptor.

Subsequently, the active compound 1, inactive compound 2 and the proposed compound X1
were docked into the ligand-binding pocket of the protein (code PDB: 5UH5), as described in
the figure 6,7B)and 8 . The results of docking shows unfavorable interactions for compound2
(the inactive molecule), while compound 1 shows tree types of interactions: pi-pi and pi-alkyl
interactions withPHE: 218 and ILE: 220residues.Furthmore van der waals interactions with
TYR:906, SER: 219,900, ASN: 902. While the proposed x1 compound shows Pi-alkyl and pi-
sulfur interactions with TYR: 906 residues and van der-waals interactions with VAL:
217,226,274 residues, GLY: 229, 275, PHE: 218,261, GLU:259, SER: 900, THR: 228, MET:
260, ASN: 273,902, ASP: 216 residues..These interactions explain the stability and the high
activity of the proposed compound.

Fig. 6: Docking interactions of compound 16 as the inactive compound in database.

43
Fig. 7: Docking interactions of compound 16 as active compound in database.

Fig. 8: Docking interactions of the proposed compound X1.

44
4. CONCLUSION:

3D-QSAR and surflex-docking methods were used to explore the structure-activity

relationship of series of 1,2,4triazole as antibacterial. The excellent predictive ability of
CoMFA and CoMSIA observed for the test set of compounds indicate that these models could
successfully be used to predict the MIC values. Furthermore, the CoMFA and CoMSIA
contour maps results offered enough information to understand the structure-activity
relationship and identified structural features influencing the activity. Thereforth, new
triazoles derivatives were proposedrelying on structure-activity relationship obtained, these
predicted compounds show excellent results of external validation and total scoring of
surflex-docking, and thus they can be trustworthy.

REFERENCES
[1]: Mohammad Asif, A Mini Review on Antimalarial Activities of Biologically Active Substituted
Triazoles Derivatives International Journal of Advanced Research in Chemical Science. 2014;1(6):22-
28
[2]: Yan S, Liu Y, Chen Y, Liu L, Lin J, An efficient one-pot synthesis of heterocycle- fused 1, 2, 3-
triazole derivatives as anticancer agents. Bio.
[3]: Banday AH, Shameem SA, Gupta BD, Kumar HMS. D-ring substituted 1,2,3-triazolyl 20-keto
pregnenanes as potential anticancer agents: Synthesis and biological evaluation. Steroids.
2010;75;801–4.
[4]: Barbuceanu SF, Gabriela LA, Ioana S, Constanatin D, Radu S, Barbuceanu F. New S-alkylated
1,2,4-triazoles incorporating diphenyl sulfone moieties with potential antibacterial activity. J Serb
Chem Soc. 2009;74(10):1041-9.
[5]: A.L. Barry, Procedure for testing antimicrobial agents in agar media, in V.L. Corian (Ed.),
Antibiotics in laboratory medicine, Williams and Wilkins, Baltimore, MD, 1991.
[6] : R. Davis, A. Markham, J. Balfour, An updated review of its pharmacology, therapeutic efficacy
and tolerability, Drugs, 1996, vol. 51, pp. 1019−1026.
[7]: Comparative QSAR: Toward a Deeper Understanding of Chemicobiological Interactions.Hansch
C, Hoekman D, Gao HChem Rev. 1996 May 9; 96(3):1045-1076.
[8]: Z. ROSTAMI, QSAR Modeling of Antimicrobial Activity with Some Novel 1,2,4- Triazole
Derivatives, Comparison with Experimental Study, Iranian Journal of Mathematical Chemistry, Vol.
4, No. 1, March 2013, pp. 91 – 109. DOI: 10.22052/IJMC.2013.5284.
[9]: M. Clark, R. D. Cramer, and N. Van Opdenbosch, “Validation of the general purpose tripos 5.2
force field,” J. Comput. Chem., vol. 10, no. 8, pp. 982–1012, Dec. 1989.
[10]: W. P. Purcell and J. A. Singer, “A brief review and table of semiempirical parameters used in the
Hueckel molecular orbital method,” J. Chem. Eng. Data, vol. 12, no. 2, pp. 235–246, Apr. 1967
[11]: M. D. M. AbdulHameed, A. Hamza, J. Liu, and C.-G. Zhan, “Combined 3D-QSAR Modeling
and Molecular Docking Study on Indolinone Derivatives as Inhibitors of 3-Phosphoinositide-
Dependent Protein Kinase-1,” J. Chem. Inf. Model., vol. 48, no. 9, pp. 1760–1772, Sep. 2008.
[12]: Y. Fukunishi, S. Yamasaki, I. Yasumatsu, K. Takeuchi, T. Kurosawa, H.Nakamura, Mol.
Inf. 2017, 36, 1600013.
[13]: L. Ståhle and S. Wold, “Multivariate data analysis and experimental design in biomedical
research.,” Prog. Med. Chem., vol. 25, pp. 291–338, 1988.
[14]: Bush BL, Nachbar RB Jr (1993) Sample-distance partial least squares: PLS optimized for many
variables, with application to CoMFA. J Comput Aided Mol Des 7:587–619.

45
[15]: Sybyl 8.1; Tripos Inc.: St. Louis, MO, USA, 2008; Available online: http://www.tripos.com
(accessed on 26 January 2011).
[16]: Jain, A.N. Surflex: Fully automatic flexible molecular docking using a molecular Similarity-
Based search engine. J. Med. Chem. 2003, 46, 499–511.
[17]: Sun, J.Y.; Cai, S.X.; Mei, H.; Li, J.; Yan, N.; Wang, Y.Q. Docking and 3D-QSAR study of
thiourea analogs as potent inhibitors of influenza virus neuraminidase. J. Mol. Model. 2010. 16,1809–
1818.
[18]: Ai, Y.; Wang S.T.; Sun, P.H.; Song F.J. Molecular Modeling Studies of 4,5-Dihydro-
1Hpyrazolo[4,3-h] quinazoline Derivatives as Potent CDK2/Cyclin A Inhibitors Using 3D-QSAR and
Docking . Int. J. Mol. Sci. 2010, 11, 3705–3724.
[19]: Lan, P.; Chen W.N.; Chen W.M. Molecular modeling studies on imidazo[4,5-b]pyridine
derivatives as Aurora A kinase inhibitors using 3D-QSAR and docking approaches. Eur. J. Med.
Chem. 2011, 46, 77–94.
[20]: Lan, P.; Chen, W.N.; Xiao, G.K.; Sun, P.H.; Chen, W.M. 3D-QSAR and docking studies on
pyrazolo[4,3-h]quinazoline-3-carboxamides as cyclin-dependent kinase 2 (CDK2) inhibitors. Bioorg.
Med. Chem. Lett. 2010, 20, 6764–6772.

A. GHALEB et al. RHAZES: Green and Applied Chemistry Vol.3, 12, 2018, pp. 33~46

46