Molecular Docking Workflow with AutoDock Vina and ChimeraX

Molecular docking is an in silico method used to predict how and how strongly a ligand binds to the binding site of a target protein. It is widely used in drug discovery and drug design (1). This tutorial explains the basic preparation steps for molecular docking using UCSF ChimeraX and the docking procedure performed with AutoDock Vina on an M1 MacBook Air.

Requirements for the Docking Procedure

Docking simulations can be performed using AutoDock Vina, which is compatible with all major operating systems, including Windows 10 or later, macOS (10.15 Catalina or later), and Linux. For molecular structure preparation and visualization, auxiliary software such as UCSF Chimera or ChimeraX (version 1.8 or later) is commonly employed (2-4).

Step 1: In this tutorial, the crystal structure of the COVID-19 main protease in complex with the inhibitor N3 (PDB ID: 6LU7) is used as the target protein. After being retrieved from the Protein Data Bank (PDB), the target protein is saved in PDB format and opened in ChimeraX, as shown in Figure 1.

Figure 1. Screenshot of the retrieved structure in ChimeraX.

Step 2: To accurately identify the active site, it is essential to identify the existing inhibitor. To do this, choose the inhibitor by clicking on Select > Chains > N-[(5-METHYLISOXAZOL-3-YL)CARBONYL]ALANYLL-VALYL-N~1~-((1R,2Z)-4-(BENZYLOXY)-4-OXO-1-{[(3R)-2-OXOPYRROLIDIN-3- YL]METHYL}BUT-2-ENYL)-L-LEUCINAMIDE, as shown in Figure 2.

Figure 2. Screenshot of selection.

After the selection, the chains must be accorded to a color. To differentiate the selected residues from the remainder of the protein (Figure 3.), modify the color by navigating to Actions > Color >  purple (or any preferred color).

Figure 3. Screenshot of differentiating the selected residues.

Step 3:  Subsequently, to optimize the receptor for docking, Dock Prep tools will be utilized. In this manner, click on Tools > Structure Editing > Dock Prep (Figure 4.).

Şekil 4. Screenshot of selecting Dock Prep tool.

Select every option in the dock preparation box, except “Delete non-complexed ions,” and then  click OK (Figure 5.).

Figure 5. Screenshot of the Dock Prep tool.

Step 4: Then, add hydrogen into the protein by choosing the relevant parameters below and clicking OK  (Figure 6.). We permit the software to choose the most suitable option based on the model by picking the previously mentioned alternatives.

Şekil 6. Adding hydrogen atoms to the protein.

As shown in Figure 7, apply Gasteiger charges to the protein by selecting the option and confirming with OK.

Figure 7. Adding Gasteiger charges delete native ligands.

Step 5: Subsequently, delete native ligands (in this case Chain C, which were colored previously) and save this file as pre_6lu7.pdb (5).

Figure 8. Delete native ligands

Molecular Docking

In molecular docking, the definition of the grid box is a critical step, as the search for potential  binding poses is restricted to this defined region. The size and coordinates of the grid box can be determined through several approaches: 

1. Based on Literature Data

Previously published docking studies often report optimized grid parameters for specific protein–ligand systems. Utilizing such values ensures comparability and reproducibility of results across studies. 

2. Using the Co-Crystallized (Native) Ligand

If the protein structure contains a co-crystallized ligand, the geometric center of this ligand can be used to define the grid box center. This method ensures that the docking is focused on the experimentally validated binding site. Visualization tools such as ChimeraX or PyMOL can be employed to calculate the centroid coordinates of the ligand. 

3.Defined by Active-Site Residues

When key catalytic or binding residues are known (e.g., His41 and Cys145 in the SARS CoV-2 main protease), the box can be centered on the geometric average of these residues’ coordinates. The box dimensions should then be adjusted according to the expected ligand size, typically ranging from 20 to 30 Å. 

4. Blind Docking Approach

In cases where no binding site information is available, a sufficiently large grid box  encompassing the entire protein can be applied. This strategy allows for unbiased  exploration of potential binding cavities, though it is computationally more demanding. 

In this tutorial, the grid box coordinates were obtained from the literature (6,7). However, as an alternative, they can be calculated using the center point of the co-crystallized ligand, the coordinates of the active site, or by applying a blind docking strategy that covers the entire protein.

Step 1: After successfully installing AutoDock Vina on your system, the next step in performing the docking procedure is to generate the PDBQT files. In this step, a PDBQT file of the target protein will be created, including polar hydrogen atoms and their corresponding partial charges. For this process, the mk_prepare_receptor.py script will be used in the terminal.

$ mk_prepare_receptor.py -i prep_6lu7.pdb -o 6lu7_receptor -p -v --box_size 25 25 25 --box_center -11.824 14.735 74.152 

In this command, the -i option specifies the input protein structure (prep_6lu7.pdb), while the -o option defines the output prefix (6lu7_receptor). The -p and -v options enable protonation and verbose output, respectively. The search area for docking is defined using --box_size 25 25 25, which specifies a cubic box with edges of 25 Å, centered at the coordinates --box_center -11.824 14.735 74.152.

Step 2: After this process, three main output files are generated:

1. 6lu7_receptor.pdbqt – The input file containing the rigid protein required for AutoDock Vina.

2.6lu7_receptor.box.txt – A text file containing the box dimensions in Vina format.

3.6lu7_receptor.box.pdb – A PDB file that allows visualization of the defined grid box.

Similarly, a PDBQT file for the ligand must also be generated using the Meeko Python package.

Step 3: Proper preparation of the ligand is a critical step prior to molecular docking, as it ensures that the molecule is in the correct format and conformation for the docking software. In this study, the quercetin molecule was converted into the .pdbqt format compatible with AutoDock Vina. The procedure is as follows:

The three-dimensional structure of quercetin is obtained from the PubChem database in .sdf format.

The mk_prepare_ligand.py script is used to generate the .pdbqt file for the ligand.

$ mk_prepare_ligand.py -i Conformer3D_COMPOUND_CID_5280343.sdf -o quercetin.pdbqt 

Step 4: After this step, quercetin has been successfully prepared as a .pdbqt file fully compatible with AutoDock Vina for molecular docking. This ensures that the ligand has the appropriate protonation states, rotatable bonds, and atomic charges necessary for accurate binding predictions.

The molecular docking analysis can now be initiated by running the following command.

$ vina --receptor 6lu7_receptor.pdbqt --ligand quercetin.pdbqt --config 6lu7_receptor.box.txt

--exhaustiveness=32 --out 6lu7_ligand_vina_out.pdbqt

When the exhaustiveness parameter is set to 32, the results generally yield a single docked conformation at this energy level. Due to the reduced default exhaustiveness, several poses may be presented in reverse order and may exhibit less favorable energies.

Figure 9. Terminal output screen for the molecular docking.

Figure 10. Visualization of the molecular docking result using ChimeraX software.

References

1.Butt, S. S., Badshah, Y., Shabbir, M., & Rafiq, M. (2020). Molecular Docking Using Chimera and Autodock Vina Software for Nonbioinformaticians. JMIR bioinformatics and biotechnology, 1(1), e14232. https://doi.org/10.2196/14232

2.Eberhardt, J., Santos-Martins, D., Tillack, A. F., & Forli, S. (2021). AutoDock Vina 1.2.0: New Docking Methods, Expanded Force Field, and Python Bindings. Journal of chemical information and modeling, 61(8), 3891–3898. https://doi.org/10.1021/acs.jcim.1c00203

3.Trott, O., & Olson, A. J. (2010). AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. Journal of computational chemistry, 31(2), 455–461. https://doi.org/10.1002/jcc.21334

4.Meng, E. C., Goddard, T. D., Pettersen, E. F., Couch, G. S., Pearson, Z. J., Morris, J. H., & Ferrin, T. E. (2023). UCSF ChimeraX: Tools for structure building and analysis. Protein science : a publication of the Protein Society, 32(11), e4792. https://doi.org/10.1002/pro.4792

5.Bilginer, S., Gözcü, S., & Güvenalp, Z. (2022). Molecular Docking Study of Several Seconder Metabolites from Medicinal Plants as Potential Inhibitors of COVID-19 Main Protease. Turkish journal of pharmaceutical sciences, 19(4), 431–441. https://doi.org/10.4274/tjps.galenos.2021.83548

6.Yahya, E. M. (2020). Criblage des molécules inhibitrices du SARS-CoV-2 par une approche in silico. 

7.Junapudi, S., Janapati, Y. K., Uppugalla, S., Harris, T., Yaseen, M., & Latif, M. (2023). Molecular Docking Analysis of SARS-CoV-2 Inhibitor N3 (6LU7) against Selected Flavonoids and Vitamins. Coronaviruses, 4(4), 29-36. https://doi.org/10.2174/0126667975261384231010181117