A simple click by click protocol to perform docking: AutoDock 4.2 made easy for non-bioinformaticians

Abstract

Recently, bioinformatics has advanced to the level that it allows almost accurate prediction of molecular interactions that hold together a protein and a ligand in the bound state. For instance, the program AutoDock has been developed to provide a procedure for predicting the interaction of small molecules with macromolecular targets which can easily separate compounds with micromolar and nanomolar binding constants from those with millimolar binding constants and can often rank molecules with finer differences in affinity. AutoDock can be used to screen a variety of possible compounds, searching for new compounds with specific binding properties or testing a range of modifications of an existing compound. The present work is a detailed outline of the protocol to use AutoDock in a more user-friendly manner. The first step is to retrieve required Ligand and Target.pdb files from major databases. The second step is preparing PDBQT format files for Target and Ligand (Target.pdbqt, Ligand.pdbqt) and Grid and Docking Parameter file (a.gpf and a.dpf) using AutoDock 4.2. The third step is to perform molecular docking using Cygwin and finally the results are analyzed. With due confidence, this is our humble claim that a researcher with no previous background in bioinformatics research would be able to perform molecular docking using AutoDock 4.2 program by following stepwise guidelines given in this article.

Introduction

Computer-aided docking is an important tool for gaining understanding of the binding interactions between a ligand (small molecule) and its target receptor (enzyme) (Anderson, 2003[1]; Schneider, 2010[6]) and has emerged as a reliable, cost-effective and time-saving technique for the discovery of lead compounds (Walters et al., 1998[8]; Schneider and Böhm, 2002[7]; Waszkowycz et al., 2001[10]). In recent years, the virtual screening approach for docking small molecules into a known protein structure is a powerful tool for drug design and has become an integral part of the drug discovery process. Computational tools like AutoDock offer the advantage of delivering new drug candidates more quickly and at a lower cost (Gilbert, 2004[2]; Warren et al., 2006[9]). AutoDock is an excellent non-commercial docking program that is widely used. Further, it employs a stochastic Lamarckian genetic algorithm for computing ligand conformations and simultaneously minimizing its scoring function which approximates the thermodynamic stability of the ligand bound to the target protein (Morris et al., 1998[4], 2009[5]). The use of complementary experimental and informatics techniques increases the chance of success in many stages of the discovery process. Theoretically the application of AutoDock in virtual screening is constrained only by the chemical compounds features that can be calculated and the relation between these features and the target (Lazarova, 2008[3]). But the problem arises in practical implementation of AutoDock in virtual screening of compounds which requires several considerations. Thus, this paper provides an easier protocol for the use of AutoDock for molecular docking purposes and will hopefully help in practically implementing AutoDock and AutoDock tools for the virtual screening purposes. To make it easier to understand, an example of experiment of the docking of Imipenem-hydrolyzing enzyme beta-lactamase SME-1 with Imipenem as ligand was made using AutoDock 4.2/ADT.

Requirements

1. Windows XP or Windows 7

Freely available software’s for non-commercial uses:

1. MGL tools

   http://mgltools.scripps.edu/downloads

2. Cygwin

   http://www.cygwin.com/install.html

   (Click setup-x86.exe for 32-bits version while setup-x86_64.exe for 64-bits version)

3. Discovery Studio Visualizer

   http://accelrys.com/products/discovery-studio/visualization-download.php

4. Binary files

   http://autodock.scripps.edu/downloads/autodock-registration/autodock-4-0-1-and-autogrid-4-0.0

   (Fig. 1)

   Download and Extract autodocksuite-4.0.1-i86Cygwin.tar

   Copy autodock4.exe and autogrid4.exe

   (Fig. 2)

   Paste in My computer\ C drive\ Cygwin\ bin

5. Java

   http://www.java.com/en/download/index.jsp

Figure 1. Screenshot 1.

Figure 2. Screenshot 2.

Methods

1 Retrieving Required Ligand and Target .pdb files from major databases:

1.1 Retrieving Target.pdb files from major protein databases

http://www.rcsb.org/pdb/home/home.do

(Fig. 3)

Figure 3. Screenshot 3.

• Type the query protein or enzyme (Imipenem-hydrolyzing beta-lactamase SME-1)

• Select enzyme (Imipenem-hydrolyzing beta-lactamase SME-1)

(Fig. 4)

Figure 4. Screenshot 4.

• Select download files

• Click PDB file (gz) and download it

(Fig. 5)

Figure 5. Screenshot 5.

• Open it in Discovery Studio Visualiser

• Save as .pdb format

(Fig. 6)

Figure 6. Screenshot 6.

• Press Control+H

• Select Hetatm and Delete

(Fig. 7)

Figure 7. Screenshot 7.

• Select B chain and Delete

(As both A and B chain are similar and Imipenemcan bind to anyone of the two chains)

(Fig. 8)

Figure 8. Screenshot 8.

Save as Target.pdb

(Fig. 9)

Figure 9. Screenshot 9.

1.2 Retrieving Ligand.pdb files from major ligand databases

http://www.drugbank.ca/ or

http://pubchem.ncbi.nlm.nih.gov/

• Search your Ligand (Imipenem)

(Fig. 10)

Figure 10. Screenshot 10.

• Click on Ligand (Imipenem)

(Fig. 11)

Figure 11. Screenshot 11.

• Click 3D image

• Open SDF

• Save 3D SDF

(Fig. 12)

Figure 12. Screenshot 12.

• Open 3D SDF file of Ligand in Discovery Studio visualiser

• Right Click to ‘show structure in 3D window’

(Fig. 13)

Figure 13. Screenshot 13.

• Click on 3D image and Save as Ligand.pdb file

(Fig. 14)

Figure 14. Screenshot 14.

2 Preparing PDBQT format for Target and ligand (Target.pdbqt, Ligand.pdbqt), Grid and Docking Parameter file (a.gpf and a.dpf) using AutoDock 4.2

• Open AutoDock present on desktop

  (*Created after successful installation of MGL Tools)

(Fig. 15)

Figure 15. Screenshot 15.

• Select AutoDock 4.2

• Dismiss

(Fig. 16)

Figure 16. Screenshot 16.

2.1 Preparation of Target.pdbqt file

• Open File

• Read Molecule

• Select and Open Target.pdb (*Created in first step)

(Fig. 17)

Figure 17. Screenshot 17.

• Target molecule will appear on screen

• Click on Edit

• Click on Hydrogens

• Click on Add

(Fig. 18)

Figure 18. Screenshot 18.

Click Polar Only

Click OK

(Fig. 19)

Figure 19. Screenshot 19.

• Again Edit

• Click Charges

• Add Kollman Charges

• Click OK

• Open Grid

• Click on Macromolecules

• Click on Choose

• Click Target

• Click Select Molecule

• Click OK

(Fig. 20)

Figure 20. Screenshot 20.

• Open My computer

• Open C drive

• Open Cygwin

• Open home

• Create new folder and rename it as 1 (or any other shortname)

• Save Target in Folder 1

(*In short: save Target.pdbqt in C:\Cygwin\home\1 and after saving macromolecule gets coloured)

(Fig. 21)

Figure 21. Screenshot 21.

2.2 Preparation of Ligand.pdbqt file

• Open Ligand

• Click Input

• Click Open

• Change format from .pdbqt to .pdb

(Fig. 22)

Figure 22. Screenshot 22.

• Select Ligand

• Click Open

• Click OK

• Again Open Ligand

• Click Torsion Tree

• Click Detect Root

• Again Open Ligand

• Click Torsion Tree

• Click Set Number of Torsions

• Set number of active torsions between 1 to 6

• Click Dismiss

(Fig. 23)

Figure 23. Screenshot 23.

• Again Open Ligand

• Click Aromatic Carbons

• Click Aromaticity criterion

• Click OK (* If ‘Enter angle in Degrees: 7.5’)

(Fig. 24)

Figure 24. Screenshot 24.

• Again Open Ligand

• Click Output

• Click Save as PDBQT

• Save Ligand file in C:\Cygwin\home\1

  (* In the same folder and in same way as Target.pdbqt file)

(Fig. 25)

Figure 25. Screenshot 25.

2.3 Preparation of Grid Parameter File (a.gpf)

• Open Grid

• Click Set Map Types

• Click Choose Ligand

• Click Ligand

• Click Select Ligand

(Fig. 26)

Figure 26. Screenshot 26.

• Again Open Grid

• Click Grid Box

(*We have used X,Y,Z dimension as 60x60x60. Further X,Y,Zcenter (Center Grid Box) can be changed according to the requirements but we are taking them as Default)

• Click File

• Click Close saving current

(Fig. 27)

Figure 27. Screenshot 27.

• Again Open Grid

• Click Output

• Click Save GPF

• Name the File name as a.gpf

• Save a.gpf file (.gpf format) in C:\Cygwin\home\1 (* In the same file where Target and Ligand .pdbqt files were saved)

(Fig. 28)

Figure 28. Screenshot 28.

2.4 Preparation of Docking Parameter File (a.dpf)

• Open Docking

• Click Macromolecules

• Click Set Rigid Filename

• Go to C:\ Cygwin\ home\ 1

• Select Target.pdbqt

• Click Open

(Fig. 29)

Figure 29. Screenshot 29.

• Again Docking

• Click Ligand

• Click Choose

• Click Ligand

• Click Select Ligand

(Fig. 30)

Figure 30. Screenshot 30.

• Click Accept

(Fig. 31)

Figure 31. Screenshot 31.

• Again Docking

• Click Search Parameters

• Click Genetic Algorithm

• Click Accept (*Using Default but we can change no. of GA runs)

• Again Docking

• Click Docking parameters

• Click Accept (*Using Default)

• Again docking

• Click Output

• Click LamarkianGA(4.2)

• Name the File name as a.dpf

• Save a.dpf file (.dpf format) in C:\Cygwin\home\1

  (* In the same file where Target and Ligand .pdbqtand a.gpffiles were saved)

(Fig. 32)

Figure 32. Screenshot 32.

At last four files Target.pdbqt, Ligand.pdbqt, a.gpf and a.dpf are present in the C:\ Cygwin\ home\1

(Fig. 33)

Figure 33. Screenshot 33.

3 Using Cygwin for Molecular Docking

Open Cygwin (*By clicking icon on the desktop)

Use these commands highlighted in brown font color by copy and paste in Cygwin and press enter after each command:

(cd..)cd<space>..

(ls)ls<space>

(cd 1) cd<space>1(or foldername)<space>

(ls)ls<space>

(autogrid4.exe -p a.gpf -l a.glg &)

autogrid(tab)<space>-p<space>a.gpf<space>-l<space>a.glg&

(Fig. 34)

Figure 34. Screenshot 34.

(tail -f a.glg &) tail<space>-f<space>a.glg<space>&

(Fig. 35)

Figure 35. Screenshot 35.

(autodock4.exe -p a.dpf -l a.dlg &)

autodock(tab)<space>-p<space>a.dpf<space>-l<space>a.dlg&

(tail -f a.dlg &) tail<space>-f<space>a.dlg<space>&

(Fig. 36)

Figure 36. Screenshot 36.

(After Successful Completion)

(Fig. 37)

Figure 37. Screenshot 37.

Copy Target.pdb file in C:\Cygwin\ home\1

(Fig. 38)

Figure 38. Screenshot 38.

Copy and Paste the following commands in Cygwin Window and press enter after each command:

(grep '^DOCKED' a.dlg | cut -c9- >a.pdbqt)

(cut -c-66 a.pdbqt> a.pdb)

(catTarget.pdb a.pdb | grep -v '^END ' | grep -v '^END$' > complex.pdb)

(Fig. 39)

Figure 39. Screenshot 39.

• Close Cygwin Window

• Click OK

4 Analyzing results and Retrieving Ligand-Enzyme interaction complex .pdb

4.1 Analyzing Results

• Open AutoDock

• Click Analyze

• Click Docking

• Click Open

• Select a.dlg

• Click Open

(Fig. 40)

Figure 40. Screenshot 40.

• Click OK

• Again Analyze

• Click Conformations

• Click Play

• Click &

• Click show information

• Click this sign ► to observe each conformation from 1 to 10

Note the confirmation showing best down binding energy and inhibition constant

(*In our case 10 conformation was best with binding energy (ΔG) as -5.75 and inhibition constant (Ki) as 60.87 µM)

(Fig. 41)

Figure 41. Screenshot 41.

4.2 Retrieving Ligand-Enzyme interaction complex .pdb

• Open C drive

• Open Cygwin

• Open home

• Open 1

• Open complex.pdb in Discovery Studio Visualizer

(Fig. 42)

Figure 42. Screenshot 42.

Select all other complexes and delete them except the best

(*In our case Complex model 10 was best as conformation 10 was showing best results in our case).

(Fig. 43)

Figure 43. Screenshot 43.

• Click Scripts

• Click Ligand Interactions

• Click Show Ligand Binding Site Atoms

(Fig. 44)

Figure 44. Screenshot 44.

• Right Click on Complex

• Click Label

• Select Object: AminoAcid

• Select Attributes: 1 Letter & ID insertion code

• Click OK

(Fig. 45)

Figure 45. Screenshot 45.

• Save as Image files

(Fig. 46)

Figure 46. Screenshot 46.

Conclusion

AutoDock is a popular non-commercial docking program that docks a ligand to its target protein and performs well (accurate and computationally fast). In this paper we propose an easier user-friendly docking protocol for docking ligands with target protein that utilizes AutoDock and Cygwin for docking operations. Our protocol provides a detailed outline and advice for use of AutoDock, AutoDock Tools, its graphical interface and to analyze interaction complexes using computational docking. The example of a docking experiment between Imipenem-hydrolyzing beta-lactamase SME-1 (an enzyme) and Imipenem (a ligand) using AutoDock 4.2/ADT has been given. Our sincere aim is to spread knowledge and make scientific research accessible to researchers who could not afford to buy software or pay high subscription fees of online docking servers. With due confidence, this is our humble claim that a researcher with no previous background in bioinformatics research would be able to perform molecular docking using AutoDock 4.2 program by following stepwise guidelines given in this article.

Acknowledgements

The authors are thankful to all the scientists of this world who possess a burning desire to share their knowledge and skills with the entire world free of charge and solely for the benefit of mankind and expect its reward from Allah alone. We extend sincere thanks to the inventors of ‘AutoDock’.