The three-dimensional structure of the receptor i.e. the RNA dependent RNA polymerase docked with remdesivir was obtained from the Protein Data Bank [6] (PDB id: 7BV2). Ligand structures were obtained from Zinc15 database [7]. For this study, filters like “in-vivo” and “world” filters were used to choose the ligands. Repurposing an already approved drug is faster than creating any new novel compound. Therefore the “world” approved filter was used to sort only those compounds which have been approved by a competent authority in any part of the world and not just the United States Food and Drug Administration.
Step 2: Determining active site amino acid residues in the receptor
W. Yin et al. stated in their study [8] that “The complex structure reveals that the partial double-stranded RNA template is inserted into the central channel of the RdRp where Remdesivir is covalently incorporated into the primer strand at the first replicated base pair and terminates chain elongation.” To find suitable alternatives to remdesivir, we would have to produce the same interactions that it produces with the other suitable compounds. Such type of attempt in-silico would have either produced unsatisfactory results or results which would prove to be useless in-vivo. Therefore, I figured out an alternate way to inhibit the action of RdRp. If we can inhibit the binding of the template RNA strand to RdRp then there would no production of the primer strand by replication. Even if it would occur, remdesivir could bind to it can further weaken or inhibit the process. So, the amino acids targeted in this study were those which bound to these RNA strands according to W. Yin et al. These amino acids are listed with their positions as follows:
- Y915
-
Y595
-
F594
-
S592
-
G590
-
A580
-
D684
-
A558
-
G683
-
G559
-
S682
-
K500
-
N534
-
S501
-
Q541
-
N507
-
L854
-
I847
-
R858
-
S861
-
D865
-
R836
-
A840
-
Q815
-
C813
-
S814
-
D761
-
S759
-
D760
Step 3: Preparation of receptor and ligand for docking
The receptor i.e. the RdRp protein was first processed in Drug Discovery studio [9] by removal of heterogenous atoms, water molecules, prime and template RNA strands and remdesivir. This gave us the clean RdRp molecule. Using PyRx docking software [10], it was converted automatically into an Autodock macromolecule. Around 750 drug molecules were loaded into PyRx using Open Babel plugin [11] which were then converted into Autodock ligands by minimization of their energies, addition of hydrogen atoms and addition of partial charges.
Step 4: Docking of ligands to the receptor within restricted search space of target amino acids
Using Autodock Vina [12], ligands were docked into the restricted search space containing the target amino acids in step 2. The search parameters set were as follows:
center_x = 81.0674935587
center_y = 90.6082924987
center_z = 112.717699441
size_x = 27.5120174204
size_y = 40.1660106465
size_z = 39.5156154165
The docking was conducted with maximum exhaustiveness of 4 modes. The study was conducted on a Windows 10 64-bit operating system which took about 8 hours to complete.
Step 5: Sorting of result on basis of binding affinity, interactions, inclusion and exclusion criteria.
The results obtained were filtered to include only the best mode of each ligand which had RMSD value of 0. They were then sorted from the highest to lowest order of binding affinity. The names of the ligands were derived from the Zinc15 database.
Step 6: Final potential drug candidates obtained.
Literature search about any possible association of these drugs to COVID-19 was done. Top 20 drugs were chosen and were included in this study.