Open Access | Peer-reviewed | Research Article

Azadeh Kordzadeh

Chemical and Petroleum Engineering Department, Sharif University of Technology, Tehran, Iran.

Ahmad Ramazani Saadatabadi*

Chemical and Petroleum Engineering Department, Sharif University of Technology, Tehran, Iran.

Published: December 24, 2021 DOI: 10.5281/zenodo.xxxxxxx

Abstract

The interactions between the spike protein of SARS-CoV-2 and Remdesivir, Favipiravir, Lopinavir and Hydroxychloroquine including Lennard-Jones and electrostatic potentials were investigated using molecular dynamics (MD) simulation in order to evaluate their potential capabilities for the treatment of coronavirus respiratory disease. The structural changes in the spike protein after drug adsorption were evaluated by the root mean square deviation (RMSD) and radius of gyration. The obtained results showed that Remdesivir and Hydroxychloroquine have strong interactions with the spike protein and have changed the structure of the spike protein significantly, where Favipiravir and Lopinavir showed only weak interactions. Molecular docking also confirmed that Remdesivir and Hydroxychloroquine have strong binding energy with the spike protein. Although Hydroxychloroquine has strong interaction with the spike protein, this drug did not adsorb on the active site of the spike protein which could interact with the ACE2 receptor. Contrary to Hydroxychloroquine, Remdesivir adsorbed on the active site of spike protein which confirmed the clinical reports on the effectiveness of Remdesivir comparing to Hydroxychloroquine. The steered molecular dynamics (SMD) revealed that adsorption of drug molecules with potential application in the treatment of SARS-CoV-2 should change the spike protein structure irreversibly and decrease the binding affinity between spike protein and ACE2 receptor significantly.
Keywords: SARS-CoV-2, Molecular Dynamics Simulation, Remdesivir , Favipiravir, Lopinavir , Hydroxychloroquine.
Citation: Azadeh Kordzadeh & Ahmad Ramazani Saadatabadi   (2021) Blocking mechanism of SARS-CoV-2 spike protein with Remdesivir, Favipiravir, Lopinavir and Hydroxychloroquine - A Molecular Dynamics Simulation Study, Journal of PeerScientist 4(2): e1000036.
Received: October 21, 2021; Accepted: December 18, 2021; Published: December 24, 2021.
Copyright:© 2021 Azadeh Kordzadeh & Ahmad Ramazani Saadatabadi, This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Competing interests: The authors have declared that no competing interests exist.
* E-mail: ramazani@sharif.edu | Phone: +982166165431.

Introduction

Novel coronavirus (CoV), severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was first identified in December 2019 in Wuhan with symptoms like dry cough, sore throat, nasal congestion, tiredness, fever, loss of taste, and smell which may change with a mutation in the virus [1-2]. The SARS-CoV-2, SARS-CoV and Middle East respiratory syndrome CoV (MERS-CoV) come under genus Betacorona virus [3]. The pathogen was soon identified as a pandemic across the globe so that based on World Health Organization (WHO) reports 266,504,411 cases were recorded globally on 8 December 2021 [4].

Figure 1 shows a general schematic of SARS-CoV-2 and its structural proteins. The four structural proteins of βcoronavirus are membrane (M), envelope (E), spike (S), and nucleocapsid (N) protein, mediation of coronavirus host infection is established by spike (S) protein. An investigation by scientists in China revealed that the SARS-CoV-2 requires an angiotensin-converting enzyme 2 (ACE2) receptor for their binding and invasion of the host [5-6]. Researchers are trying to find specific treatments for the new virus and several therapeutic agents have been investigated. Theoretical methods such as molecular docking and molecular dynamics not only accelerate the process of discovering a new drug but also reduce research costs. By emerging the crystal structure of the main protease of SARS-CoV-2 in protein data bank [7] many studies calculated binding energy between drug molecules and the main protease of SARS-CoV-2 to evaluate a drug candidate [8-12]. Molecular docking could give a first approximation of molecular interactions when there are several drug molecules and could be done as the first step of computational studies. Further steps give an accurate vision about molecular interactions that could be obtained with MD simulation which is a time-dependent method.

In some studies, the inhibitory of natural substances and phytocompounds against the SARS-Cov-2 have been evaluated [13-14]. Alzami et al., [15] investigated four natural compounds for blocking human ACE2 receptor. Their results showed that Andrographolide and Pterostilbene displayed better inhibition potential for ACE2. The potential of saffron which contains crocin, crocetin, picrocrocin and safranal, for treatment of SARS-CoV-2 was investigated [16-17]. Our previous study demonstrated that crocetin has a high affinity towards the spike protein and main protease of SARS-CoV-2 [18-19]. The molecular docking of withaferin-A derivative molecules exhibited a binding affinity of - 7.84 kcal/mol with the main protease of SARS-CoV-2 [20].

Figure 01: Schematic representation of SARS-CoV-2 and its structural proteins.

Also, the drugs for other diseases such as HIV, Hepatitis and Ebola were investigated on SARS-CoV-2. A molecular docking calculation indicates that anti-hepatitis C virus drugs such as Sofosbuvir, Ribavirin, and Remdesivir have high binding energy towards the main protease of SARS-CoV-2 [21]. Although the SARS-CoV-2 has a 79.6% sequence identity to SARS-CoV [22], it has been shown that SARS-CoV-2 active sites are different from SARS, so the drugs used for SARS such as ribavirin, interferon, Lopinavir -ritonavir and corticosteroids are not necessarily effective for SARS-CoV-2 [23]. WHO reports and clinical studies suggested that Hydroxychloroquine, Lopinavir /ritonavir, Remdesivir and dexamethasone could be a drug candidate for SARS-CoV-2. Investigating the effect of adding functional groups to drugs is a solution that can change the function of the drug A theoretical study shows that when Hydroxychloroquine is functionalized with the hydroxypyrrolidine group, its binding energy with the main protease of SARS-CoV-2 increases from -5.41 to -7.62 kcal/mol [24].

Most theoretical researches use the main protease of SARS-CoV-2 for finding an effective drug [25-26], while a precise look at virus structure (Figure 1) indicates that when drug molecules achieved the virus will encounter spike protein and lipid bilayer as a barrier against drug to enter the virus, so the blocking of the spike protein is a rational way to change the functionality of SARS-CoV-2. Several experimental and clinical trials investigated the effects of different drugs on SARS-CoV-2 including Remdesivir , Favipiravir, Lopinavir and Hydroxychloroquine, but the effectiveness and molecular mechanism of these drugs remain controversial, yet. To find a relation between molecular structure and effectiveness of the drugs on SARS-CoV-2 treatment, the interaction between Remdesivir, Favipiravir, Lopinavir and Hydroxychloroquine (Figure 02) with spike protein of SARS-CoV-2 were simulated by MD simulation in GROMACS software. In addition, the protonation of amine groups of Favipiravir and Hydroxychloroquine and the adsorption of these drug molecules on the spike protein and their effects on protein structure were investigated. Finally, the potential between the changed structure of the spike protein and ACE2 receptor was evaluated by steered molecular dynamics (SMD) simulation to relate binding affinity to potential treatment ability.

Figure 02:. The chemical structure of a) Remdesivir b) Favipiravir c) Lopinavir and d) Hydroxychloroquine.

Results & Discussion

Interaction Evaluation

Potential of Interaction

The binding energies between the spike protein and drug molecules which are calculated by molecular docking are illustrated in Figure 3. The results demonstrate that Remdesivir and Hydroxychloroquine have stronger binding energy toward the spike protein comparing to Lopinavir and Favipiravir. The initial and final snapshots of the simulation box for interaction between four drug molecules and the spike protein are presented in Figures 4,5,6,7. Where at beginning, the drug molecules are distributed randomly around the spike protein in simulation box, during simulation time the drug molecules approached and adsorbed on the spike protein due to Lennard-Jones (LJ) and electrostatic potentials.

Figure 03: Binding energies between the spike protein and drug molecules.

Figure 04: The snapshots of a) initial and b) final step of simulation for interaction spike protein with favipiravir. The cyan, white, pink and red vdW spheres show the carbon, hydrogen, fluorine and oxygen atoms in drug molecules, respectively. The spike protein is shown with white, green, red and blue cartoon which represent nonpolar, polar, acidic and basic amino acids, respectively. See also video S1 in Supporting Information (SI).

Figure 05: The snapshots of a) initial and b) final step of simulation for interaction spike protein with hydroxychloroquine. The color scheme is the same as Figure 4. See also video S2 in Supporting Information (SI).

Figure 06: The snapshots of a) initial and b) final step of simulation for interaction spike protein with lopinavir. The color scheme is the same as Figure 4. See also video S3 in Supporting Information (SI).

Figure 07: The snapshots of a) initial and b) final step of simulation for interaction spike protein with remdesivir. The color scheme is the same as Figure 4. See also video S4 in Supporting Information (SI).

The LJ potentials between drug molecules and spike protein are illustrated in Figure 8 that demonstrated REM has strong LJ potential with the spike protein whereas favipiravir (FAV) has weak LJ potential with the spike protein. The electrostatic potential between spike protein and drug molecules is depicted in Figure 9 which implies that remdesivir (REM) and lopinavir (LOP) have strong and weak electrostatic potential with spike protein, respectively, which could be due to different functional groups and surface charges. The equilibrium values of LJ and electrostatic potentials between drug molecules and the spike protein are shown in Table 1. The obtained results in Table 1 demonstrated that for FAV and HYD with positive charge the electrostatic potential is larger than LJ potential while the LJ potential is dominant for LOP and REM which are neutral. the Figure 5 (b) shows that at the end of simulation, five hydroxychloroquine (HYD) molecules adsorbed on spike protein. The repulsive electrostatic potential between HYD molecules which have 2 positive charges caused them to remain distributed in the bulk solution. The adsorption of lopinavir (LOP) molecules on spike protein is presented in Figure 6. This figure shows that all LOP molecules are adsorbed on the spike protein and formed agglomeration which could be due to neutral charge of LOP molecules [27]. Adsorption of remdesivir (REM) molecules on spike protein is presented in Figure 7. At the end of simulation time, still half of REM molecules are unabsorbed on the spike protein.

Figure 08: Lennard-Jones (LJ) potential energy profiles between the spike protein remdesivir (REM), favipiravir (FAV), lopinavir (LOP) and hydroxychloroquine (HYD).

Figure 09: Electrostatic potential energy profiles between the spike protein remdesivir (REM), favipiravir (FAV), lopinavir (LOP) and hydroxychloroquine (HYD).

Table 01: The equilibrium Lennard-Jones (LJ) and electrostatic (ELE) potential energies (kJ/mol) between the spike protein and four drug molecules:

Interaction sites

The interaction sites between spike protein and ACE2 receptor [28] are illustrated in Figure 10(a), which named active sites, because it could interact with ACE2 receptor. So, if the drug molecule blocks the active sites of spike protein, it could prevent adsorption of spike protein on ACE2 receptor.

Figure 10: a) The spike protein structure in water as a reference system. The dotted line shows the interaction sites of the spike protein with ACE2 receptor. Adsorption sites of b) Favipiravir, c) Hydroxychloroquine, d) Lopinavir, and e) Remdesivir on the spike protein. The color scheme is same as Figure 4. See also video S5 in Supporting Information (SI).

The final snapshot at 100 ns as illustrated in Figure 10 (b) reveals that seven molecules of favipiravir (FAV) have adsorbed on the spike protein and three of them still remained unabsorbed. From presented results in Figure 10 (b), it could be concluded that there is no agglomeration between drug molecules which should be due to repulsive electrostatic potential between FAV molecules having positive charge. As, it is illustrated in Figure 10 (b) FAV molecules with positive charge are adsorbed on aspartic acid (ASP427), a negatively charged amino acid, and threonine (THR523-THR415-THR333), cysteine (CYS336) and asparagine (ASN334-ASN394) as polar amino acids. Comparison between Figures 10 (b) and 10 (a) indicates that FAV adsorbed on active site of spike protein at appropriate situation. Figure 10 (c) demonstrates that HYD molecules adsorbed on polar amino acids serine (SER373) and threonine (THR526). It seems that the HYD molecules with positive charge due to attractive electrostatic forces adsorbed on aspartic acid (ASP427) which is a negatively charged amino acid and similarly, FAV has adsorbed on ASP427. Also, the spike protein has two positive charges (at pH=7) and so the repulsive electrostatic force with positively charged FAV and HYD prevented some drug molecules to adsorb on the spike protein. Comparison between Figure 10 (c) and 10 (a) indicates that HYD molecules have not adsorbed on active site of the spike protein. Figure 10 (d) indicates that LOP molecules are adsorbed on Tyrosine (TYR473-TYR489) and Phenylalanine (PHE338) aromatic amino acids of the spike protein. Comparison between Figure 10 (d) and 10 (a) implies that LOP molecules interact with active site of spike protein. Figure 10 (e) demonstrates that REM molecules interact with serine (SER469-SER477-SER479-SER375-SER373-SER381) which is a polar amino acid. Comparison between Figure 10 (e) and 10 (a) reveals that REM molecules have adsorbed on active site of the spike protein.The comparison of Figure 10(c) and 10(e) indicates that HYD and REM have similar adsorption sites on ASP427 and SER373.

Radial Probability

The radial probability of finding drug molecules around the spike protein is illustrated in Figure 11. As it is depicted in Figure 11, HYD and REM molecules adsorbed at radius of 1.412 and 1.652 from the center of mass of the spike protein. Short distances of these two drug molecules confirmed strong electrostatic and LJ potential between drug molecules and the spike protein that were discussed in potential of interactions section.

Figure 11: The radial distribution function (RDF) of remdesivir (REM), favipiravir (FAV), lopinavir (LOP) and Hydroxychloroquine (HYD) referenced to center of mass of the spike protein.

Conformational Changes

RMSD

To investigate the effects of the drug adsorption on protein structure, the Root Mean Square Deviation (RMSD) was evaluated as it is illustrated in Figure 12. The equilibrium values of RMSD in water medium and after drug adsorption are shown in Table 2. As it is presented in Figure 12 and Table 2, the RMSD of the spike protein after adsorption of LOP is similar to water medium without drug molecules which is the reference system. It can be concluded that LOP adsorption has not changed the spike protein conformation significantly. The RMSD of the spike protein after FAV adsorption reached to 1.2959 nm. The change of RMSD after adsorption of REM and HYD is remarkable. The RMSD of the spike protein after REM and HYD adsorption is larger than 3 nm which confirms the strong potential energies with the spike protein in potential of interactions section. The low fluctuation of RMSD curve in each system proves that all systems are in equilibrium and simulation time was sufficient. The magnitude of RMSD is in accordance with previous studies [29-30].

Figure 12: The root mean square deviation of spike protein in water medium and after adsorption of REM, LOP, HYD and FAV.

Table 02: The equilibrium root mean square deviation (RMSD) and radius of gyration (Rg) of the spike protein in water medium and after drug adsorption:

Radius of Gyration

The radius of gyration of spike protein in water medium considered as reference system and the radius of gyration of spike protein after drug adsorption was compared with that of the reference system as it is presented in Figure 13. The equilibrium values of radius of gyration are reported in Table 2. The radius of gyrations of spike protein after adsorption of REM and HYD significantly changed which also confirm RMSD results. After adsorption of REM and HYD the radius of gyration of the spike protein increased respectively to 1.8729 nm and 1.8571 nm. The increment of the radius of gyration indicates the expansion of the residues (amino acids) around the center of mass, therefore confirms the conformational changes of the spike protein.

Figure 13: The radius of gyration of spike protein after drug adsorption and in water medium without drug (reference system).

Figure 14: Radial probability of water molecules around the spike protein.

Hydrophilicity

The radial probability of finding water molecules around the spike protein in different simulated systems are illustrated in Figure 14 where two peaks reveal two layers of water molecules around the spike protein. Decreasing the radius of first peak after drug adsorption confirms increasing hydrophilicity of the spike protein. The number of hydrogen bonds between the spike protein and drug molecules and solvent accessible surface area (SASA) of the spike protein are reported in Table 3. The results in Tables 1 clarify that the drug adsorption have changed the structure of the spike protein and increased the hydrophilicity of it which is in accordance with decreasing radius of first layer of water molecules around the spike protein in Figure 14.

Table 03: Number of hydrogen bond (NHB) between the spike protein and water and solvent accessible surface area (SASA) of the spike protein in different simulated systems:

SMD simulation

The profile of the pulling force of spike protein towards ACE2 receptor versus displacement is depicted in Figure 15. At beginning, the spike protein is placed at a distance of 7nm from ACE2 receptor (zero displacement) and at this situation the interactions between spike protein and ACE2 receptor are weak. By pulling the spike protein toward ACE2, the LJ and electrostatic interactions increased and result in increment in pulling forces. The maximum point in force-displacement profiles identifies the binding energy between the spike protein and the ACE2 receptor. The maximum binding energy between ACE2 receptor and spike protein was observed in water medium without drug molecule. When the drug molecules adsorbed on spike protein its structure changes as discussed in sections RMSD and Radius of Gyration. Figure 15 reveals that the spike protein after drug adsorption has lower affinity toward ACE2 receptor and the pulling force decreased from 3300 pN to 1400 pN.

Figure 15: The force-displacement profile of spike protein during binding to ACE2 receptor.

Conclusion

To find a mechanism of effective potential drugs on the treatment of SARS-CoV-2 pandemic sickness, the molecular simulation of adsorption of remdesivir, favipiravir, lopinavir and hydroxychloroquine on spike protein of SARS-CoV-2 were investigated. The LJ and electrostatic interactions and adsorption sites of drug molecules were considered to evaluate drug molecule effectiveness. Presented results showed that not only, the effective drug molecule should have strong LJ and electrostatic interaction with spike protein, but also should interact with active site of spike protein to block the binding of spike protein to ACE receptor. The structural changes of spike protein after drug adsorption and its effects on receptor binding were evaluated. The between several drug molecules and spike protein calculated. The radial distribution function was used to determine the distribution of drug molecules around the spike protein. The RMSD and radius of gyration were calculated to evaluate the change in protein structure. The molecular dynamics (MD) and molecular docking results showed that FAV and LOP molecules did not have strong interactions with spike protein which confirms the binding energies which calculated by molecular docking and so they could not be effective drugs for treatment of SARS-CoV-2 which also confirmed clinical trials of these drugs. Obtained results depicted that due to weak LJ and electrostatic interactions, the structural changes of spike protein after LOP and FAV adsorption were not remarkable where REM and HYD with strong interactions with spike protein could remarkably change structure of the spike protein. It is also found that adsorption sites of these two drug molecules on the spike protein are different which should be the main reason for effectiveness of REM comparing to HYD with even strong interaction forces towards spike protein. Actually, it seems that the adsorption site of REM on the spike protein decreases affinity of binding spike protein to ACE2 receptor comparing to HYD. The obtained results of this work could certainly help design new more effective drugs with ability of blocking SARS-CoV-2 virus for potential treatment of SARS-CoV-2  pandemic sickness.

Materials & Methods

The Structural Model

The initial structure of remdesivir, favipiravir, lopinavir and hydroxychloroquine were obtained from DRUGBANK server [31]. The spike protein of SARS-CoV-2 with PDB ID 6M0J was obtained from protein data bank server [7].

Molecular Docking Study

To calculate the binding energy between the spike protein and drug molecules, molecular docking study was performed using Autodock Vina 4.2 [32] with a grid box of dimension 76 Å × 92 Å × 160 Å with grid spacing 1 Åand 120 runs using the Lamarckian genetic algorithm (LGA).

Force Field

The GROMACS 5.1.4 simulation package [33] was employed to perform all simulations. The visual molecular dynamics (VMD 1.9.1) program [34] was used for molecular visualization. The all-atom GROMOS54A7 force field was employed to calculate all bonded and nonbonded interactions [35]. The simple point charge [36] model was selected for water molecules. The surface charges of drug molecules were calculated by MarvinSketch 18.10 software [37]. At pH=7.4 the charge of remdesivir and lopinavir is zero and the charge of hydroxychloroquine and favipiravir is positive 2 and 1, respectively. Topology files of drug molecules were obtained using PRODRG server [38] and the partial charges are to be corrected by performing the corresponding quantum mechanics QM calculations. Quantum mechanics (QM) calculations performed by GAMESS software [39] employing DFT 6-31G(d,p) B3LYP method [40-41], along with monitoring variation of molecular conformation to achieve minimum energy of molecules. The partial charges of the atoms were calculated using electrostatic potential method (ESP) [42] and considering this fact that the molecules are present in water medium.

MD Simulation

The simulations were performed in the NPT ensemble. The Nose´-Hoover thermostat was used to keep ensemble temperature at 37 (physiological temperature) [43]. The Parrinello–Rahman barostat [44] with semi-isotropic pressure coupling algorithm [45], maintains the system pressure at 1 bar and also, the LINCS algorithm [46] was used to constrain all bonds. The particle mesh Ewald method was employed for calculation of the long-range electrostatic interaction with a cut off radius of 1.2 nm [47]. The Lennard-Jones potential was used to calculate the van der Waals (vdW) interaction with a cut off radius of 1.2 nm as recommended [48]. To clarify the effect of drug molecules, at first, the spike protein was simulated in water medium without drug molecules which was selected as the reference system. Then, for remdesivir, favipiravir, lopinavir and hydroxychloroquine, 10 drug molecules were added into the previous simulation box for each system.The simulation box dimensions for each system were (). The NVT simulation for 10 ns was performed and temperature was set at natural body temperature. The pressure was adjusted at atmospheric pressure in NPT ensemble for 10 ns. Then the MD step was performed for 100 ns for data collection and last 30 ns were used for analysis.

For investigation of the effects of the protein structure on the binding energy with ACE2 receptor a steered molecular dynamics (SMD) simulation was performed. The spike protein was placed at a distance of 7 nm from the center of mass of ACE2 in a simulation box with dimension of (). The NVT simulation for 10 ns was performed at same condition used for the MD simulation. Then, the constant velocity SMD was performed. In which the spike protein was attached to dummy atoms via a virtual spring and was moved at a constant velocity. The force needed for the displacement of the dummy atoms to an imaginary point had been calculated using the following equations [49]:

Where,  is the potential energy gradient, k is the spring force constant, v is the velocity of pulling which could be supposed to be equal to 0.01  , t is the current time, r is the instantaneous vector position, r0 is the initial vector position of the SMD atom, and n is the vector direction which the dummy atom is pulled.

Authors’ contribution: A.R. Saadatabadi: Conceived and designed the experiments; Analyzed and interpreted the data.
A. Kordzadeh: Performed the experiments; wrote the paper.

References

  1. Bogoch, Isaac I., et al. "Pneumonia of unknown aetiology in Wuhan, China: potential for international spread via commercial air travel." Journal of travel medicine2 (2020): taaa008. 
  2. Helmy, Yosra A., et al. "The COVID-19 pandemic: a comprehensive review of taxonomy, genetics, epidemiology, diagnosis, treatment, and control." Journal of clinical medicine4 (2020): 1225.
  3. Woo, Patrick CY, et al. "Coronavirus genomics and bioinformatics analysis." viruses8 (2010): 1804-1820.
  4. World Health Organization. Available online: https://covid19.who.int/
  5. Stadler, Konrad, et al. "SARS—beginning to understand a new virus." Nature Reviews Microbiology3 (2003): 209-218.
  6. Zhou, Peng, et al. "Discovery of a novel coronavirus associated with the recent pneumonia outbreak in humans and its potential bat origin." BioRxiv(2020).
  7. Burley, Stephen K., et al. "RCSB Protein Data Bank: biological macromolecular structures enabling research and education in fundamental biology, biomedicine, biotechnology and energy." Nucleic acids researchD1 (2019): D464-D474.
  8. Adhikari, Puja, et al. "Intra-and intermolecular atomic-scale interactions in the receptor binding domain of SARS-CoV-2 spike protein: implication for ACE2 receptor binding." Physical Chemistry Chemical Physics33 (2020): 18272-18283.
  9. Owis, Asmaa I., et al. "Molecular docking reveals the potential of Salvadora persica flavonoids to inhibit COVID-19 virus main protease." RSC Advances33 (2020): 19570-19575.
  10. Procacci, Piero, et al. "Interaction of hydroxychloroquine with SARS-CoV2 functional proteins using all-atoms non-equilibrium alchemical simulations." Chemical Communications62 (2020): 8854-8856.
  11. Sethi, Aaftaab, et al. "Understanding the role of galectin inhibitors as potential candidates for SARS-CoV-2 spike protein: in silico studies." RSC Advances50 (2020): 29873-29884.
  12. Świderek, Katarzyna, and Vicent Moliner. "Revealing the molecular mechanisms of proteolysis of SARS-CoV-2 M pro by QM/MM computational methods." Chemical Science39 (2020): 10626-10630.
  13. Bello, Martiniano, Alberto Martínez-Muñoz, and Irving Balbuena-Rebolledo. "Identification of saquinavir as a potent inhibitor of dimeric SARS-CoV2 main protease through MM/GBSA." Journal of molecular modeling12 (2020): 1-11.
  14. Motwalli, Olaa, and Meshari Alazmi. "Analysis of natural compounds against the activity of SARS-CoV-2 NSP15 protein towards an effective treatment against COVID-19: a theoretical and computational biology approach." Journal of molecular modeling6 (2021): 1-11.
  15. Alazmi, Meshari, and Olaa Motwalli. "Molecular basis for drug repurposing to study the interface of the S protein in SARS-CoV-2 and human ACE2 through docking, characterization, and molecular dynamics for natural drug candidates." Journal of molecular modeling12 (2020): 1-10.
  16. Aanouz, I., et al. "Moroccan Medicinal plants as inhibitors against SARS-CoV-2 main protease: Computational investigations." Journal of Biomolecular Structure and Dynamics8 (2021): 2971-2979.
  17. Husaini, Amjad M., Khan Nadiya Jan, and Gowher A. Wani. "Saffron: A potential drug-supplement for severe acute respiratory syndrome coronavirus (COVID) management." Heliyon(2021): e07068.
  18. Kordzadeh, Azadeh, and Ahmad Ramazani Saadatabadi. "Effects of the Temperature and the pH on the Main Protease of SARS-CoV-2: A Molecular Dynamics Simulation Study." (2021).
  19. Kordzadeh, Azadeh, Ahmad Ramazani Saadatabadi, and Amin Hadi. "Investigation on penetration of saffron components through lipid bilayer bound to spike protein of SARS-CoV-2 using steered molecular dynamics simulation." Heliyon12 (2020): e05681.
  20. Ghosh, Arabinda, et al. "Structure-activity relationship (SAR) and molecular dynamics study of withaferin-A fragment derivatives as potential therapeutic lead against main protease (M pro) of SARS-CoV-2." Journal of molecular modeling3 (2021): 1-17.
  21. Elfiky, Abdo A. "Ribavirin, Remdesivir, Sofosbuvir, Galidesivir, and Tenofovir against SARS-CoV-2 RNA dependent RNA polymerase (RdRp): A molecular docking study." Life sciences253 (2020): 117592.
  22. Li, Fang. "Structure, function, and evolution of coronavirus spike proteins." Annual review of virology3 (2016): 237-261.
  23. Bzówka, Maria, et al. "Molecular dynamics simulations indicate the COVID-19 Mpro is not a viable target for small-molecule inhibitors design." BioRxiv(2020).
  24. Mishra, Deepak, et al. "Structurally modified compounds of hydroxychloroquine, remdesivir and tetrahydrocannabinol against main protease of SARS-CoV-2, a possible hope for COVID-19: Docking and molecular dynamics simulation studies." Journal of Molecular Liquids335 (2021): 116185.
  25. Kiriwan, Duangnapa, and Kiattawee Choowongkomon. "In silico structural elucidation of the rabies RNA-dependent RNA polymerase (RdRp) toward the identification of potential rabies virus inhibitors." Journal of molecular modeling6 (2021): 1-10.
  26. Paul, Debarati, Debadrita Basu, and Shubhra Ghosh Dastidar. "Multi-conformation representation of Mpro identifies promising candidates for drug repurposing against COVID-19." Journal of molecular modeling5 (2021): 1-16.
  27. Kordzadeh, Azadeh, et al. "Adsorption and encapsulation of the drug doxorubicin on covalent functionalized carbon nanotubes: A scrutinized study by using molecular dynamics simulation and quantum mechanics calculation." Journal of Molecular Graphics and Modelling88 (2019): 11-22.
  28. Lan, Jun, et al. "Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor." Nature7807 (2020): 215-220.
  29. Manandhar, Suman, et al. "Structure-based docking, pharmacokinetic evaluation, and molecular dynamics-guided evaluation of traditional formulation against SARS-CoV-2 spike protein receptor bind domain and ACE2 receptor complex." Chemical Papers(2021): 1-21.
  30. Yun, Yangfang, et al. "Identification of therapeutic drugs against COVID-19 through computational investigation on drug repurposing and structural modification." Journal of Biomedical Research6 (2020): 458.
  31. Wishart, David S., et al. "DrugBank 5.0: a major update to the DrugBank database for 2018." Nucleic acids researchD1 (2018): D1074-D1082.
  32. Trott, Oleg, and Arthur J. Olson. "AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading." Journal of computational chemistry2 (2010): 455-461.
  33. Van Der Spoel, David, et al. "GROMACS: fast, flexible, and free." Journal of computational chemistry16 (2005): 1701-1718.
  34. Humphrey, William, Andrew Dalke, and Klaus Schulten. "VMD: visual molecular dynamics." Journal of molecular graphics1 (1996): 33-38.
  35. Schmid, Nathan, et al. "Definition and testing of the GROMOS force-field versions 54A7 and 54B7." European biophysics journal7 (2011): 843-856.
  36. Berendsen, Herman JC, et al. "Interaction models for water in relation to protein hydration." Intermolecular forces. Springer, Dordrecht, 1981. 331-342.
  37. Csizmadia, Péter. "MarvinSketch and MarvinView: molecule applets for the World Wide Web." (1999).
  38. Van Aalten, Daan MF, et al. "PRODRG, a program for generating molecular topologies and unique molecular descriptors from coordinates of small molecules." Journal of computer-aided molecular design3 (1996): 255-262.
  39. Bode, Brett M., and Mark S. Gordon. "MacMolPlt: a graphical user interface for GAMESS." Journal of Molecular Graphics and Modelling3 (1998): 133-138.
  40. Becke, Axel D. "Density‐functional thermochemistry. I. The effect of the exchange‐only gradient correction." The Journal of chemical physics3 (1992): 2155-2160.
  41. Schmidt, Michael W., et al. "General atomic and molecular electronic structure system." Journal of computational chemistry11 (1993): 1347-1363.
  42. Smith, Daniel GA, et al. "Quantum Chemistry Common Driver and Databases (QCDB) and Quantum Chemistry Engine (QCEngine): Automation and interoperability among computational chemistry programs." The Journal of Chemical Physics20 (2021): 204801.
  43. Evans, Denis J., and Brad Lee Holian. "The nose–hoover thermostat." The Journal of chemical physics8 (1985): 4069-4074.
  44. Rühle, Victor. "Pressure coupling/barostats." Journal Club(2008).
  45. Yousefpour, Abbas, et al. "Molecular dynamics simulation of nonsteroidal antiinflammatory drugs, naproxen and relafen, in a lipid bilayer membrane." International Journal of Quantum Chemistry15 (2013): 1919-1930.
  46. Zheng, Hui-dong, Bi-yu Wang, and Yan-xiang Wu. "Molecular dynamics simulation on the interfacial features of phenol extraction by TBP/dodecane in water." Computational and Theoretical Chemistry1-3 (2011): 66-72.
  47. Darden, Tom, Darrin York, and Lee Pedersen. "Particle mesh Ewald: An N⋅ log (N) method for Ewald sums in large systems." The Journal of chemical physics12 (1993): 10089-10092.
  48. Sutmann, G.J.Q.s.o.c.m.-b.s.f.t.t.a. Classical molecular dynamics. 10 (2002): 211-254.
  49. Isralewitz, Barry, Mu Gao, and Klaus Schulten. "Steered molecular dynamics and mechanical functions of proteins." Current opinion in structural biology 11.2 (2001): 224-230.