Structure Prediction and In-Silico Designing of Drugs against Fork head-Box R2 Protein

 

Dr. Manish Devgan

Professor, Faculty of Pharmacy, R.P. Educational Trust Group of Institutions, Bastara, Karnal-132001, Haryana, India.

*Corresponding Author E-mail: manishdevgan12@gmail.com

 

ABSTRACT:

Forkhead-box (FOX) family proteins, involved in cell growth and differentiation as well as embryogenesis and longevity, are DNA-binding proteins regulating transcription and DNA repair.

Objective: FOXR2 promotes cell proliferation and malignancy in hepatocellular carcinoma (HCC) and could be a novel promising therapeutic target for this disease.

Methods: In this work, an in-silico model of FOXR2 protein was generated using the approach of homology modeling and loop modeling. The model was validated with Ramachandran plot analysis. The ligands were generated with the help of Drug bank, ZINC data base, and Kegg data base and were docked against FOXR2 protein using online server Patchdock. The structure of ligand ZINC 3830634 with the maximum score was varied by using ACD/ChemSketch 8.0 and the docking was done for the resulting 10 new ligands.

Results and Conclusion: The results indicated that the designed ligands 3 and 10 shows better docking score than that showed by ZINC 3830634. Thus, justifies further studies needed for the development of potent inhibitors for the over expression of FOXR2 protein making the management of HCC more efficient.

 

KEYWORDS: Hepatocellular carcinoma; FOXR2 protein; docking; molecular modelling; homology modeling.

 


INTRODUCTION:

Hepatocellular carcinoma (HCC) is the fifth most prevalent cancer in men and eighth most prevalent cancer in women worldwide, resulting in at least 5 lakhs deaths per year. It is responsible for 90 % of all liver cancers [1]. Its iteration is considered to be high (>20 cases/100,000 inhabitants/year) in the far East and Africa, medium (5 to 20 cases/100,000) inhabitants/year) in Europe, and low (<5 cases/100,000 inhabitants/year) in South America [2]. HCC exhibits the main complexity of cirrhosis, and shows an increasing appearance worldwide related to the increased existence of the various risk factors of chronic liver diseases, such as hepatitis infection with hepatitis C and B viruses, and more recently Fatty Liver Diseases which are correlated with metabolic syndrome.

 

Although most HCC develop in the background of chronic liver disease, some may occur on normal liver and usually relates to specific types, including fibrolamellar HCC mostly detected young population, or malignant transformation of hepatocellular adenomas [3]. The other risk factors for HCC include alcoholism, aflatoxin, nonalcoholic steatohepatits (if progression to cirrhosis has occurred), hemochromatosis, Wilson’s disease, type 2 diabetes and hemophilia. HCC may present with yellow skin, bloating from fluid in the abdomen, easy bruising from blood clotting abnormalities, loss of appetite, unintentional weight loss, abdominal pain especially in the right upper quadrant, nausea, vomiting, or feeling tired [4]. Tumors are multifocal in the liver in 75% of cases at diagnosis. Diagnosis is usually made by history, physical examination, imaging (ultrasound, MRI or CT scan showing a liver mass consistent with HCC) and optionally elevated serum α-fetoprotein (AFP), i.e., >400 ng/ml, because AFP is elevated in only 50%-75% of cases [1]. The main predictive factors of HCC are linked to tumor stage (number and size of nodules, presence of vascular invasion and extrahepatic spread), liver function (defined by Child-Pugh’s class, bilirubin, albumin, portal hypertension) and general health status [3]. The treatment plan should be established on the presence or absence of liver cirrhosis, extent of disease, growth pattern of tumor, hepatic functional reserve and patient’s performance status. The relevant treatment feasibilities include surgical (liver resection, liver transplantation), ablative (transarterial chemoembolization, radiofrequency ablation) and medical (sorafenib) modalities [1].  Forkhead-box (FOX) family proteins, engaged in cell growth and differentiation as well as embryogenesis and longevity, are DNA-binding proteins managing transcription and DNA repair [5]. FOX gene family consists of at least 43 members, including FOXA1-3, FOXB1, FOXC1-2, FOXD1-6, FOXE1-3, FOXF1-2, FOXG1, FOXH1, FOXI1, FOXJ1-3, FOXK1-2, FOXL1-2, FOXM1, FOXN1-6, FOXO1-4, FOXP1-4 and FOXQ1. FOXE3-FOXD2 (1p33), FOXQ1-FOXF2-FOXC1 (6p25.3), and FOXF1-FOXC2-FOXL1 (16q24.1) loci are FOX gene clusters within the human genome. Members of FOX subfamilies A-G, I-L and Q are grouped into class 1 FOX protein, while members of FOX subfamilies H and M-P were grouped into class 2 FOX proteins. FOX super family genes are involved in carcinogenesis via gene amplification, retroviral integration and chromosomal translocation [6,7]. One study demonstrated that peculiarly expressed FOX genes and their downstream targets are engaged in the pathogenesis of Hodgkin lymphoma through deregulation of B-cell differentiation and may symbolize as a useful diagnostic markers and/or therapeutic targets [8]. Another study concluded that pituitary tumor transforming gene (PTTG1) is a FOXM1 targeted gene. FOXM1 binds to PTTG1 promoter to enhance PTTG1 transcription, and FOXM1-PTTG1 pathway bolsters colorectal cancer migration and invasion [9]. One study functionally confirmed FOXR2 as proto-oncogene engaged in malignant peripheral nerve sheath tumors (MPNSTs) sustenance [10]. Another study implicated FOXR2 as potential therapeutic target for medulloblastomas as well as for cancers that involve Sonic Hedgehog pathway signaling [11]. Another study identified FOXR2 as an important molecular marker for diagnosis and prognosis of breast cancer [12]. Another study suggested that FOXR2 promotes cell proliferation and malignancy in HCC and could be a novel promising therapeutic target for this disease [13]. All these studies indicate that FOXR2 protein should be comprehensively investigated to develop novel therapeutics and preventives for human diseases. Three-dimensional protein structures are precious sources of information for the functional interpretation of protein molecules. Yet, determining experimental structures of many proteins presents technical challenges. These structures are best determined by experimental methods such as X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy. However, these methods require pure preparations of proteins at concentration much higher than those at which the proteins exist in a physiological environment. Due to these reasons, atomic structures of many medicinal and biologically important proteins do not exist. In such cases, prediction of the protein structure by computational methods can ordinarily result in a beneficial model. Comparative modeling, or homology modeling, exploits the fact that two proteins whose sequences are evolutionarily connected display identical structural appearance. Thus, the acknowledged structure of a protein (template) can be used to accomplish a molecular model of the protein (query) whose experimental structure is not known [14]. Computer Aided Drug Designing is swiftly becoming an important tool in drug discovery, the in-silico study has provided the realization of the interaction between receptor and ligands [15-18]. In this study, the structure of FOXR2 protein was designed by using homology modeling. The docking of the ligands was done to predict the binding orientation of small drug molecules with their protein target in order to prognosticate the affinity and activity of the small molecules in inhibiting FOXR2 protein so that it may lead to undermine the proliferation and migratory ability of HCC cells.

 

MATERIALS AND METHODS:

The hardware used for calculating molecular modeling includes a personal computer with Intel (R) Core (TM) i3 CPU processor, Windows 7 Home Premium 32-bit operating system having RAM of 2.00 GB.

 

Sequence Alignment:

Fast alignment (FASTA)

The FASTA format is a text based format for depicting either nucleotide sequences or peptide sequences, in which nucleotides or amino acids are depicted using single letter codes. A sequence in FASTA format begins with a single line narration, followed by lines of sequence data. The description line is demarcated from the sequence data by a greater-than (“>”) symbol in the first column [19]. The FASTA sequence of FOXR2 was obtained from the website of National Centre for Biotechnology Information [20]. Basic Local Alignment Search Tool (BLAST) The BLAST is an algorithm for evaluating primary biological sequence information, such as the amino acid sequence of different proteins or the nucleotides of DNA sequences [21]. The FASTA was used and standard protein BLAST was executed on the NCBI. The BLAST-P was obtained using protein data bank proteins data base [22]. 

 

Three Dimensional Position-Specific Scoring Matrix: (3D-PSSM) The 3D-PSSM is a fast web based technique for protein fold identification using 1D and 3D sequence profiles paired with secondary structure and solvation potential information. The FASTA sequence was submitted to 3D-PSSM for fold recognition [23, 24].

 

Protein Homology/Analogy Recognition Engine (Phyre):

Phyre2 is a suite of tools available on the web to predict and evaluate protein structure, function and mutations [25]. The FASTA sequence was submitted to Phyre for amino acid sequence prediction [26].

 

Templates Preparation:

The data obtained from BLAST, 3D-PSSM and Phyre was evaluated at the RCSB protein data bank. The Protein Data Bank (PDB) archive is the single worldwide archive of the 3D structures of hefty biological molecules, including proteins and nucleic acids [27]. The templates were selected on the ground of their resolution (Å) and R-value. All the above templates were obtained by X-ray crystallography method in PDB.

 

Molecular Modeling:

Homology modeling of FOXR2 was done by using Easy Modeller. Easy Modeller is a front-end graphical interface to Modeller developed using Perl/Tk, which can be used as a standalone tool in windows platform with Modeller and Python preinstalled. Easy Modeller can produce 3-D structural models of proteins from sequence and given template(s) information using Modeller in backend [28]. The Swiss-Pdb viewer, an application that supports a user friendly interface allowing analysis of several proteins at the same time, was installed [29].

 

Structure Prediction:

The chosen six templates were submitted to the Easy Modeller. All the fifteen prepared models were analyzed based on DOPE, Molpdf and GA341 methods. The Discreet Optimized Protein Energy (DOPE) score is a statistical tool to evaluate homology models in protein structure prediction. In the Modeller objective function (molpdf), the Easy Modeller minimizes the objective function $ F$ with respect to Cartesian coordinates of ~ 10,000 atoms (3D points) that form a system (one or more molecules). The GA341 method uses the percentage sequence identity between the template and the model as a parameter. The model with the minimum molpdf and DOPE score, and the GA341 value lying in between zero and one (the higher the better) can be chosen as the best possible model [30].

 

Validation of Predicted Model:

The validation of all the fifteen models was performed by submitting the PDB files to Rampage for Ramachandran plot assessment. Rampage is a program for visualising and assessing the Ramachandran plot of a protein structure. On the basis of a set of high-quality protein structures (from the Richardson's Group at Duke University) and a number of filters (such as B-factor cutoff and van der Waals clashes), reference phi/psi plots were derived for Gly, Pro, pre-Pro and general (other) residue types, and subdivided into "favoured", "allowed" and "outlier" regions [31]. The Ramachandran plot validated the result.

 

Loop Modeling:

The protein function is determined by its shape and the physiochemical properties of its exposed surface, thus it is important to design an explicit model for protein/ligand interaction studies. The co-ordinate file was submitted for loop optimization to ModLoop, a web server for automated modeling of loops in protein structures. The server counts on the loop modeling routine in MODELLER that anticipates the loop conformations by satisfaction of spatial restraints, without depending upon a database of known protein structures. This structure was evaluated by Ramachandran plot using Rampage. The process of loop modeling and successive validation was carried on until an optimized structured model of protein was accomplished [32, 33].

 

Ligand Generation:

Various databases such as DrugBank, Kegg Data Base, Zinc database, Google Scholar, etc. were utilized to search for the chemical entities that would interact with the FOX family proteins.

 

Molecular Docking:

Docking is a method which anticipates the most fitting orientation of one molecule to a second when bound to each other to form a stable complex. Molecular docking helps in studying drug/ligand or receptor/protein interactions by identifying the suitable active sites in protein, obtaining the best geometry of ligand-receptor complex and calculating the energy of interaction for different ligands to design more effective ligands. One key aspect of molecular docking is calculating the energy of conformations and interactions using methods ranging from quantum mechanics to purely empirical energy functions. Molecular docking energy evaluations are usually carried out with the help of a scoring function [34]. The macromolecule and the ligands were prepared for docking by using Pymol and ChemBio3D software [35, 36]. The molecular docking was done against FOXR2 protein using an online server Patchdock. The Patchdock is an algorithm for molecular docking. It is inspired by object recognition and image segmentation techniques used in computer vision. The algorithm has three main stages: a) Molecular shape representation; b) Surface patch matching; c) Filtering and scoring. The input is two molecules of any type: proteins, DNA, peptides, drugs. The output is a list of potential complexes sorted by shape complementarity criteria [37, 38]. The prominent compound was preferred based on scoring.

 

Ligand Designing and Docking:

The chosen ligand was employed to design novel molecules with the help of ACD/ChemSketch 8.0 freeware. The Lipinski’s rule of five was used to check the hypothetical effectiveness of the drugs. These structures were subjected to ChemBio3D for energy minimization. The molecular docking of these sketched molecules was done against the FOXR2 protein by using Patchdock.

 

RESULTS AND DISCUSSION:

Template Generation:

The NCBI was employed to secure FASTA sequence of FOXR2 protein. The GenBank No. is AAH12934.1 and gi no. is 34189747. It is a 311 amino acid protein. The BLAST was performed on the NCBI and 05 hits were recorded as shown in Figure 1.

Figure 1 Distribution of 05 BLAST hits on the query sequence (query Id: qi/34189747/qb/AAH12934.1) in pdb protein database and the program is BLASTP 2.3.1 +.

The 3D-PSSM and Phyre were utilized for prediction of protein structure. The information received from BLAST, 3D-PSSM and Phyre was evaluated at the RCSB protein data bank. The obtained results were arranged in the descending order of % ID followed by ascending order of Resolution as shown in Table 1.

 

 

 

Table 1 Generation of templates using Blast, 3D-PSSM, Phyre and RCSB protein data bank.

S. No

Template/ Accession No

ID %

Resolution

(Ǻ)

R-Value (Obs/

Free)

1

2C6Y

32

2.40

0.258

2

3L2C

29

1.87

0.228

3

3G73

28

2.21

0.234

4

3CO7

27

2.91

0.276

5

1HST

24

2.6

0.270

6

4I1L

16

2.1

0.289

 

 

The six templates (2C6Y, 4I1L, 3G73, 3CO7, 1HST, and 3L2C) were preferred on the ground of their chains, ID %, resolution (≤ 3 Å) and the R-value (≤ 0.5). A total of fifteen models were generated with the help of Easy Modeller (Table 2). Models with the lowest DOPE assessment score and Molpdf or with the highest GA341 assessment score have the most stable minimized energy. The models number 11 shows least DOPE score and more GA341 score as compared to model number 1 and 7. Moreover, Molpdf score of model number 11 is nearly equal to that of model number 1 and 7. Thus, model number 11 was selected on these bases for further analysis.


 

Table 2 DOPE score and Ramachandran plot analysis of the fifteen possible models of FOXR2 protein.

S.No

Model No.

Molpdf

DOPE

GA341

Residues in favoured regions

(Number/

Percentage)

Residues in allowed regions

(Number/

Percentage)

Residues in outlier regions

(Number/

Percentage)

1

B99990001

6191.33691

-15958.14453

0.00705

293/94.8

10/3.2

6/1.9

2

B99990002

6379.89746

-15439.59668

0.00686

285/92.2

19/6.1

5/1.6

3

B99990003

6282.25586

-15883.35059

0.00510

288/93.2

13/4.2

8/2.6

4

B99990004

6221.36377

-15988.53809

0.00243

281/90.9

22/7.1

6/1.9

5

B99990005

6414.95166

-16004.89160

0.00497

279/90.3

23/7.4

7/2.3

6

B99990006

6249.56836

-14986.18750

0.01151

292/94.5

10/3.2

7/2.3

7

B99990007

6189.18018

-15301.92383

0.00811

288/93.2

16/5.2

5/1.6

8

B99990008

6315.87012

-16195.24414

0.00760

287/92.9

18/5.8

4/1.3

9

B99990009

6223.13721

-16290.04492

0.00423

294/95.1

10/3.2

5/1.6

10

B99990010

6294.93359

-16184.93555

0.00350

297/96.1

6/1.9

6/1.9

11

B99990011

6194.07715

-16508.15625

0.00990

293/94.8

12/3.9

4/1.3

12

B99990012

6334.94775

-16366.78809

0.00209

281/90.9

22/7.1

6/1.9

13

B99990013

6258.54199

-15335.58887

0.00235

290/93.9

15/4.9

4/1.3

14

B99990014

6394.80615

-15263.62500

0.00473

290/93.9

11/3.6

8/2.6

15

B99990015

6402.18750

-14741.81641

0.00538

288/93.2

14/4.5

7/2.3

 


Validation:

The models were further validated by Ramachandran plot, by submitting the files to Rampage. The models number 8, 11, and 13 are having equal percentage of residues in combined favoured and allowed regions. However, out of these three models, only model number 11 shows maximum percentage of residues in the favoured region (Table 2).

 

Loop modeling:

The PDB file format of model number 11 was endorsed for loop optimization to ModLoop and the output model was evaluated with the help of Ramachandran plot obtained using Rampage as well as PDBsum. As per Rampage, the protein model is having maximum percentage (99.4 %) of residues in favoured region and 0.6 % residues in allowed regions with no residues in outlier region. As per PDBsum, the protein model is having maximum percentage (92.9 %) of residues in most favoured region and 7.1 % residues in additional allowed regions with no residues in generously allowed as well as disallowed regions (Figure 2 and Table 3). The model of FOXR2 protein (Figure 3) was successfully submitted to Protein model data base (http://bioinformatics.cineca.it/PMDB/) bearing the PMDB ID: PM0080460.


 


 

Figure 3 Optimized model of FOXR2 protein


Table 3 PROCHECK statistics.

Ramachandran Plot Statistics*

G-Factors**

Regions

No. of Residues

Percentage (%)

Parameters

Score

Average Score

Most favoured regions [A, B, L]

247

92.9

Dihedral angles

Additional allowed regions [a, b, l, p]

19

7.1

Phi-psi distribution

-0.27

 

Generously allowed regions [~a, ~b, ~l, ~p]

0

0.0

chil-chi2 distribution

-0.25

 

Disallowed regions [XX]

0

0.0

chil only

0.20

 

Non-glycine and non-proline residues

266

100

chi3 and chi4

0.42

 

End residues (excl. Gly and Pro)

2

-

Omega

-0.01

 

Glycine residues

12

-

Average Score

-0.07

 

Proline residues

31

-

Main-chain covalent forces

Total no. of residues

311

-

Main-chain bond lengths

-0.24

 

 

 

 

Main-chain bond angles

-0.62

 

 

 

 

Average score

-0.46

 

 

 

 

Overall average

-0.21

 

 

Table 4 The docking results of ligands generated using Drug Bank and ZINC data base against FOXR2 protein as target.

S.No

Doxorubicin and  similar ligands

Patch Dock Score

S.No

Sorafenib & similar ligands

PatchDock Score

1

ZINC 3830631

6028

16

DB00398

5792

2

ZINC 3830632

6044

17

DB08896

5944

3

ZINC 3830633

5192

18

DB06938

5984

4

ZINC 3830634

6422

19

DB08043

5316

5

ZINC 3830729

6072

20

ZINC 3817152

5876

6

ZINC 3830730

6114

21

ZINC 6745272

5590

7

ZINC 3830731

6094

22

ZINC 16052883

5976

8

ZINC 3830732

6178

23

ZINC 37760652

4252

9

ZINC 3917708

6268

24

ZINC 38461223

5912

10

ZINC 3918087

6118

25

ZINC 40178922

5288

11

ZINC 15449300

5910

26

ZINC 48786914

4334

12

ZINC 42876243

5720

27

ZINC 59182076

5684

13

ZINC 43649326

6086

28

ZINC 61719766

4868

14

ZINC 43649328

5778

29

ZINC 89629954

5736

15

ZINC 43649330

5656

30

ZINC 91302028

5860

 

31

ZINC 49544597

4680

*Based on an analysis of 118 structures of resolution of at least 2.0 Angstroms and R-factor no greater than 20.0 a good quality model would be expected to have over 90% in the most favoured regions [A,B,L]. **G-factors provide a measure of how unusual, or out-of-the-ordinary, a property is. Values below -0.5 – unusual; Values below -1.0 - highly unusual.


 

 

Ligand Generation and Docking:

The drug doxorubicin was found to be useful in the heaptocellular carcinoma disorder [39]. Also during the literature survey the drug sorafenib was found to be effective in the management of hepatocellular carcinoma [40]. The data bases such as Drug Bank and Zinc Data base were utilized to obtain drugs similar to doxorubicin and sorafenib. These were docked against FOXR2 protein using Patchdock. The results (Table 4) indicated that the best score (6422) is given to the ligand ZINC 3830634 (Figure 4, 5). The result suggested that the compound could be a promising ligand for the target FOXR2 protein.

 

Figure 4 The structures of ligand ZINC 3830634.

 

 

 

Figure 5 Docking of ligand ZINC 3830634 with FOXR2 protein.

 

 

Ligand Designing and Docking:

The structural variation was done in the molecule ZINC 3830634 and 10 new compounds were designed with the help of ACD/ChemSketch 8.0. The docking of these compounds was done against FOXR2 protein using Patchdock. The results (Table 5, Figure 6, 7) indicated that out of all these compounds, ligand3 and ligand10  dock with the score (6716 and 6794 respectively) greater than that of the docking of the parent ligand ZINC 3830634 (6422).

Table 5 the docking results of ligands generated using chemSketch against FOXR2 as target.

Ligands

Molecular Formula

Formula Weight

Patch dock Score

ZINC 3830634

C27H30NO10

528.5272514

6422

Ligand1

C27H34N5O6

524.5882114

6118

Ligand2

C27H34N5O6

524.5882114

6250

Ligand3

C27H28Cl2NO10

597.4173714

6716

Ligand4

C27H34NO8

500.5602114

6000

Ligand5

C27H36NO6

470.5772914

5828

Ligand6

C27H36NO4

438.5784914

6234

Ligand7

C27H38NO2

408.5955714

6136

Ligand8

C27H30NO7S3

576.7240514

6182

Ligand9

C27H30NO9S

544.5928514

6034

Ligand10

C27H32NO8P2

560.4918534

6794

 

 

 

Figure 7 Ligand 10  a) Chemical structure; b) Docking pattern of ligand 10 with FOXR2 protein.

 

CONCLUSIONS:

The homology modeling and loop modeling methodology were employed to design model of FOXR2 protein. The Ramachandran plot evaluated the models. The Drug bank, ZINC data, KEGG data base were used to identify various ligands. The molecular docking done against FOXR2 protein of these ligands using online server Patchdock, identified ZINC 3830634 ligand with maximum score. The structure of this compound was varied by using ACD/ChemSketch 8.0 and then docking was done against the target protein. The present study indicates that the in silico molecular docking studies of sketched ligands, i.e., ligand 3 and 10 with FOXR2 protein manifested favorable binding interactions and justifies further studies (in vitro as well as in vivo) required for the evolution of potent inhibitors for the over expression of FOXR2 protein so as to develop a novel compound for the prevention and treatment of hepatocellular carcinoma.

 

REFERENCES:

1.     Jelic S, Sotiropoulos GC. Hepatocellular carcinoma: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2010;21 Suppl 5:V59-64.

2.     Franca AVC et al. Diagnosis, staging and treatment of hepatocellular carcinoma. Braz J Med Biol Res. 2004 Nov;37(11):1689-1705.

3.     Paradis V. Histopathology of hepatocellular carcinoma. In:   Brouquet A, editor. Multidisciplinary treatment of hepatocellular carcinoma. Heidelberg (Germany): Springer Science and Business Media; 2012. p. 21-32.

4.     Wikipedia, The free encyclopedia [Internet]. San Franscisco (CA): Wikipedia Foundation, Inc (US); 2001-2016. BLAST [updated 2016 April 18; cited 2016 April 21]; [About 17 p.]. Available from: https://en.wikipedia.org/wiki/Hepatocellular_carcinoma

5.     Katoh M, et al. Cancer genetics and genomics of human FOX family genes. Cancer Lett. 2013 Jan 28;328(2):198-206.

6.     Katoh M, Katoh M. Characterization of human FOXN4 gene in silico. Int J Mol Med. 2004 Nov;14(5):949-53.

7.     Katoh M, Katoh M. Human FOX gene family (review). Int J Oncol. 2004 Nov;25(5):1495-5000.

8.     Nagel S, et al. Deregulated FOX genes in Hodgkin lymphoma. Genes Chromosomes Cancer. 2014 Nov;53(11):917-33.

9.     Zheng Y, et al. Fox M1 transactivates PTTG1 and promotes colorectal cancer cell migration and invasion. BMC Med Genomics. 2015 Aug 12;8:49.

10.   Rahmann EP, et al. Forward genetic screen for malignant peripheral nerve sheath tumor formation identifies new genes and genetic pathways driving tumorigenesis. Nat Genet. 2013 Jul;45(7):756-66.

11.   Koso H, et al. Identification of FoxR2 as an oncogene in medulloblastoma. Cancer Res. 2014 Apr 15;74(8):2351-61.

12.   Song H, et al. High expression of FOXR2 in breast cancer correlates with poor prognosis. Tumour Biol. 2015 Nov 23:1-7. doi: 10.1007/s13277-015-4437-4. PubMed PMID: 26596836.

13.   Wang X, et al. FOXR2 contributes to cell proliferation and malignancy in human  hepatocellular carcinoma. Tumor Biol. 2016 Feb 5:1-9. doi: 10.1007/s13277-016-4923-3. PubMed PMID: 26846213.

14.   Ramachandran S, Dokholyan NV. Homology modeling: generating structural models to understand protein function and mechanism. In: Dokholyan N, editor. Computational modelling of biological systems: from molecules to pathways, biological and medical physics, biomedical engineering. New York: Springer Science+Business Media, LLC; 2012. p. 97-116.

15.   Devgan M. Homology modeling and molecular docking studies of DNA replication licensing factor minichromosome maintenance protein 5 (MCM5). AJPTech. 2015;5(1):17-22. doi: 10.5958/2231-5713.2015.00004.5 

16.   Devgan M. Structure prediction and insilico designing of drugs for the inhibition of EphA10 tyrosine kinase receptor protein. Acta Bio Sci. 2016;3(2):86-94.

17.   Devgan M, et al. In-silico designing of drugs for the inhibition of AMF-HER2 complex in trastuzumab resistant breast cancer. Indian J Biotechnol. Forthcoming 2016.

18.   Devgan M. Structure prediction and in-silico designing of drugs against homeobox C8 protein. J Chem Pharm Res. 2016 April;8(3):891-99.

19.   Wikipedia, The free encyclopedia. FASTA format [Internet]. San Franscisco (CA): Wikipedia Foundation, Inc (US); c2001- [cited 2016 April 25]. Available from: https://en.wikipedia.org/wiki/FASTA_format

20.   National Center for Biotechnology Information. Forkhead box R2 [Homo sapiens] [Internet]. Bethesda (MD): National Library of Medicine (US); [cited 2016 Mar 30]. Available from: http://www.ncbi.nlm.nih.gov/protein/34189747?report=fasta

21.   Wikipedia, The free encyclopedia. BLAST [Internet]. San Franscisco (CA): Wikipedia Foundation, Inc (US); 2001- [cited 2016 Mar 30]. Available from: https://en.wikipedia.org/wiki/BLAST

22.   National Center for Biotechnology Information. BLASTP [Internet]. Bethesda (MD): National Library of Medicine (US); [cited 2016 Mar 30]. Available from: http://blast.ncbi.nlm.nih.gov/Blast.cgi

23.   Kelly L. 3D-PSSM [Internet]. South Kensington Campus (London): Imperial College of Science, Technology and Medicine (UK); c1999 [updated 2016 Mar 30; cited 2016 Mar 30]. Available from: http://www.sbg.bio.ic.ac.uk/~3dpssm/index2.html

24.   Kelley LA, MacCallum RM, Sternberg MJ. Enhanced genome annotation using structural profiles in the program 3D-PSSM. J Mol Biol. 2000 Jun 2;299(2):499-520.

25.   Kelley LA, et al. The Phyre2 web portal for protein modeling, prediction and analysis. Nat Protoc. 2015 Jun;10(6):845-58. doi: 10.1038/nprot.2015.053.

26.   Kelly L, Sternberg M. Phyre2 [Internet].  South Kensington Campus (London): Structural Bioinformatics Group, Imperial College (UK); [cited 2016 Mar 30]. Available from: http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index 

27.   Research Collaboratory for Structural Bioinformatics. RCSB Protein Data Bank [Internet]. New Jersey, California:  Rutgers, UCSanDiego (US); [cited 2016 Mar 31]. Available from: http://www.rcsb.org/pdb/home.do

28.   Kuntal BK. EasyModeller (Version 2.0) [Internet]. Hyderabad: Kuntal Kumar Bhusan, University of Hyderabad (India); [cited 2016 Mar 31]. Available from: http://www.sites.google.com/site/bioinformatikz 

29.   Duex N, et al. DeepView/Swiss-PdbViewer [Internet]. Bazel:  Structural Bioinformatics Groups, Swiss Institute of Bioinformatics, Biozentrum University of Bazel (Switzerland); c1995-2001 [cited 2016 Mar 31]. Available from: http://Spdbv.vital-it.ch/  

30.   Procedure [Internet]. Powai (Mumbai): Department of Biosciences and Bioengineering, Indian Institute of Technology. c2015- [cited 2016 Mar 31]. Available from: http://www.bio.iitb.ac.in/~sanjeeva/virtual_lab/Virtual_Laboratory/exp-10/procedure.pdf

31.   Bakker PD, Lovell S. Rampage [Internet].  Cambridge (UK): Department of Biochemistry, University of Cambridge. [Cited 2016 Mar 5]. Available from: http://mordred.bioc.cam.ac.uk/~rapper/rampage.php 

32.   Wikipedia, The free encyclopedia [Internet]. San Franscisco (CA): Wikipedia Foundation, Inc (US). Loop modeling [updated 2015 July 13; cited 2016 April 06]; [About 4 p.]. Available from: https://en.wikipedia.org/wiki/Loop_modeling 

33.   A Fiser. ModLoop [Internet].  San Franscisco (CA): University of California (USA);   c2000 – [cited 2016 April 06]. Available from: http://modbase.compbio.ucsf.edu/modloop/  

34.   Bachwani M, Kumar R. Molecular docking: a review. IJRAP. 2011;2(6):1746-51.

35.   PerkinElmer. ChemBio3D [Internet]. Waltham, Massachusetts: PerkinElmer Inc. (USA); c1998-2016 [cited 2016 April 15]. Available from: http://www.cambridgesoft.com/Ensemble for chemistry/chemBio3D/

36.   Schrödinger. PyMol [Internet]. New York:  Schrödinger (USA); c2005-16 [cited 2016 April 15]. Available from: http://www.pymol.org  

37.   D Duhovny; R Nussinov; HJ Wolfson. Efficient unbound docking of rigid molecules. Gusfield; R Guigo, Editors. Proceedings of the 2nd Workshop on Algorithms in Bioinformatics(WABI) Rome, Italy, Lecture Notes in Computer Science 2452, Springer Verlag, 2002; 185-200,

38.   Schneidman-Duhovny D, et al. PatchDock and symmdock: servers for rigid and symmetric docking. Nucleic Acids Res. 2005 Jul 1;33(web server issue):W363-7. doi: 10.1093/nar/qki481. PubMed PMID: 15980490; PubMed Central PMCID: PMC1160241.

39.   Kyoto encyclopedia of genes and genomes. KEGG disease H00048 [Internet]. Kyoto University, Japan: Kanehisa laboratories (Japan). c1995-2016 [cited 2016 April 14]. Available from: http://www.kegg.jp/entry/ds:H00048

40.   Wikipedia, The free encyclopedia. Hepatocellular carcinoma [Internet]. San Franscisco (CA): Wikipedia Foundation, Inc (US); 2001- [cited 2016 April 21]. Available from: https://en.wikipedia.org/wiki/Hepatocellular_carcinoma

 

 

 

 

 

Received on 25.06.2017         Modified on 13.08.2017

Accepted on 09.11.2017      © RJPT All right reserved

Research J. Pharm. and Tech. 2018; 11(3): 913-920.

DOI: 10.5958/0974-360X.2018.00170.1