Structure-Guided Antiviral Design: In Silico Docking and Molecular Dynamics of SARS-CoV-2 Spike Protein Variants

Introduction

The spike glycoprotein of SARS-CoV-2 is the primary determinant of host cell entry and the principal target of neutralizing antibodies. Variants such as Delta and Omicron have accumulated mutations in the receptor-binding domain (RBD) that alter binding affinity to the angiotensin-converting enzyme 2 (ACE2) receptor and enable immune evasion [1, 2, 3]. Understanding the structural and dynamic consequences of these mutations at atomic resolution is essential for predicting zoonotic spillover risk and for rational design of antiviral agents [4, 5, 6]. Structure-guided computational approaches, including homology modeling, molecular docking, and molecular dynamics simulations, have become indispensable tools in this effort [7, 8, 9].

This article reviews the methodological pipeline for structure-guided antiviral design as applied to SARS-CoV-2 spike protein variants. Emphasis is placed on in silico techniques that predict how sequence changes alter protein-protein interactions at the RBD-ACE2 interface and at epitopes recognized by neutralizing antibodies [1, 10]. The discussion draws on parallel studies of other SARS-CoV-2 targets, such as the main protease (Mpro) and papain-like protease (PLpro), to illustrate the general principles of computer-aided drug design [11, 12, 13, 14].

Structural Biology of the Spike RBD and ACE2 Interface

The RBD of the SARS-CoV-2 spike protein adopts a five-stranded antiparallel beta-sheet core with a receptor-binding motif (RBM) that directly contacts the N-terminal helix of ACE2 [1, 6]. High-resolution crystal structures and cryo-electron microscopy maps have revealed that the RBD undergoes a hinge-like conformational transition between “up” and “down” states, with the up state being necessary for ACE2 engagement [4, 5]. Sequence surveillance data archived in the GISAID database demonstrate that mutations such as N501Y, K417N, and E484K in the RBM can increase ACE2 affinity or reduce antibody recognition [2, 10].

Homology modeling is often the first computational step when a variant structure is unavailable [15]. Using a known template structure (e.g., PDB 6M0J for the wild-type RBD-ACE2 complex), residues are mutated and the local geometry is optimized with energy minimization [16]. The accuracy of the model depends on sequence identity; for SARS-CoV-2 variants, identity exceeds 95%, making template-based modeling reliable [5, 6]. Model quality can be assessed with Ramachandran plots and MolProbity scores [16].

Molecular Docking of Variant RBD to ACE2 and Antibodies

Molecular docking predicts the preferred orientation of a ligand (e.g., ACE2 or an antibody fragment) when bound to the RBD [1, 10]. Rigid and flexible docking protocols are used. In rigid docking, both receptor and ligand are treated as static; in flexible docking, selected side chains are allowed to rotate to accommodate induced fit [16]. Common algorithms include AutoDock Vina and Glide, which use scoring functions that evaluate van der Waals, electrostatic, and desolvation contributions [15, 16].

For the RBD-ACE2 complex, docking calculations can recapitulate the crystallographic binding mode with root-mean-square deviations below 2.0 Å [1, 3]. Docking scores correlate with experimentally measured binding affinities; for example, the Omicron RBD shows a higher docking score toward ACE2 compared to the wild type, consistent with its increased transmissibility [2]. Docking is also used to predict the impact of RBD mutations on antibody binding [10]. A panel of neutralizing antibodies can be docked to variant RBD models to identify mutations that disrupt key hydrogen bonds or salt bridges, thereby enabling immune escape [4, 6].

Molecular Dynamics Simulations of the RBD-ACE2 Complex

Molecular dynamics (MD) simulations provide a time-resolved view of conformational fluctuations and binding stability. Simulations are typically run for tens to hundreds of nanoseconds using all-atom force fields (e.g., AMBER ff14SB, CHARMM36) and explicit solvent models (e.g., TIP3P) [14, 16]. The RBD-ACE2 complex is solvated in a water box, neutralized with counterions, and energy minimized before production runs. Temperature and pressure are controlled using algorithms such as Langevin dynamics and the Parrinello-Rahman barostat [14, 15].

MD trajectories are analyzed to compute root-mean-square fluctuation (RMSF), radius of gyration (Rg), and intermolecular hydrogen bond occupancy [5, 16]. Variants often exhibit altered flexibility in the RBM loop; for instance, the Omicron RBD shows reduced RMSF in regions that contact ACE2, suggesting a more stable binding interface [2, 3]. Principal component analysis can reveal collective motions relevant to conformational selection [14].

Binding Free Energy Calculations

End-point free energy methods such as Molecular Mechanics Poisson-Boltzmann Surface Area (MM-PBSA) or Generalized Born Surface Area (MM-GBSA) are routinely applied to estimate the relative binding affinity of variant complexes [1, 14]. These methods average the gas-phase molecular mechanics energy, solvation free energy, and entropic contribution over an MD trajectory. The MM-PBSA approach has been validated against experimental binding constants for RBD-ACE2 [1, 2].

Studies comparing wild-type and variant RBDs have shown that MM-PBSA computed ΔΔG values correlate well with surface plasmon resonance measurements [1, 3]. For example, the N501Y mutation enhances binding by approximately 1.5 kcal/mol through improved pi-pi stacking with ACE2 residue Y41 [2]. Similarly, E484K reduces antibody binding by 2.3 kcal/mol due to loss of a salt bridge with the antibody complementarity-determining region [10].

Structure-Guided Design of Antiviral Agents

The atomic-level insights obtained from docking and MD simulations directly inform the design of entry inhibitors, nanobodies, and peptide mimetics [1, 4, 5]. Structure-guided optimization has been a central theme in developing SARS-CoV-2 antivirals across multiple targets [7, 8, 9, 11, 17, 12, 18, 19, 20, 13, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35]. For spike, three classes of designed agents are prominent:

1. Peptide inhibitors that mimic the ACE2 interface. Ferková et al. designed proteomimetics that compete with ACE2 for RBD binding, using the structure of the RBD-ACE2 interface as a template [1]. These compounds showed nanomolar affinity in vitro.

2. Nanobodies engineered against the RBD. Hannula et al. used computational docking to select nanobody variants with enhanced neutralization breadth [4]. Jiang et al. designed trivalent nanobody clusters that simultaneously bind three RBDs, achieving avidity-driven potency [6].

3. Aptamers and other synthetic binders. Rahman et al. developed bivalent aptamers that block RBD-ACE2 interaction by bridging two RBD monomers [10].

MD simulations are used to assess the stability of designed binders and to confirm that they engage the intended epitope without inducing unfavorable conformational changes [3, 5]. The binding free energy of the designed binder to the RBD is computed with MM-PBSA and compared to that of the natural receptor [1, 2].

Predicting Immune Escape and Antigenic Drift

Combining docking of antibody panels with MD-based binding free energy calculations enables quantitative mapping of escape mutations [10]. Deep mutational scanning data can be integrated into these pipelines to prioritize mutations that simultaneously increase ACE2 affinity and reduce antibody recognition [2, 3]. This approach is instrumental for updating vaccine antigen design and for selecting monoclonal antibodies that target conserved epitopes [4, 6].

For veterinary applications, predicting how emerging variants might adapt to ACE2 orthologs of domestic animals (e.g., cats, dogs, ferrets, livestock) is crucial for risk assessment [1, 5]. Homology models of animal ACE2 can be built from the human template, and docking scores with various RBD variants can highlight potential spillover hosts [2, 3]. Such cross-species analyses rely on the same structure-guided framework described above.

Workflow Overview

The following Mermaid diagram summarizes the computational workflow for structure-guided antiviral design targeting the SARS-CoV-2 spike protein:

flowchart TD
    A[Retrieve variant sequence from GISAID], > B[Build homology model of RBD]
    B, > C[Validate model (Ramachandran, MolProbity)]
    C, > D[Prepare receptor (PDB, protonation, grid generation)]
    C, > E[Prepare ligand (ACE2 ectodomain or antibody Fv)]
    D, > F[Molecular docking (AutoDock Vina, Glide)]
    E, > F
    F, > G[Select top poses based on docking score]
    G, > H[MD equilibration (NVT, NPT) and production run]
    H, > I[Trajectory analysis (RMSF, H-bonds, PCA)]
    I, > J[MM-PBSA/GBSA binding free energy calculation]
    J, > K{Acceptable ΔG?}
    K, Yes, > L[Identify key pharmacophore features]
    K, No, > M[Refine model or mutate RBD]
    L, > N[Design inhibitor (peptide, nanobody, aptamer)]
    N, > O[Dock inhibitor to RBD]
    O, > P[MD of complex and MM-PBSA]
    P, > Q[Experimental validation]

Tables of Key Software and Methods

Conclusion

Structure-guided antiviral design that integrates homology modeling, molecular docking, and molecular dynamics simulations provides a powerful framework for understanding SARS-CoV-2 spike protein variant behavior and for developing countermeasures. The same methods successfully applied to protease and methyltransferase targets [7, 8, 9, 11, 17, 12, 19, 13, 14, 23, 24, 25, 26, 28, 29, 30, 31, 32, 33, 34, 35] are now being adapted to the spike protein, where they inform both antibody escape prediction and the rational design of entry inhibitors [1, 2, 3, 4, 5, 6, 10]. Continued integration with high-throughput sequence surveillance and deep mutational scanning will further enhance the predictive power of these in silico pipelines.

References

[1] Ferková S, Fayolle A, Boisvert O et al. Structure-Guided Design of Proteomimetics Targeting the SARS-CoV-2 S-RBD/hACE2 Interface. J Med Chem. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/42289940/

[2] Ortega Del Campo S, Fernández Ballester GJ, Blanes Mira C et al. Synergistic antiviral effects of structure-guided peptides and a mutagenic base analog on SARS-CoV-2 replication. Antimicrob Agents Chemother. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/42037394/

[3] Zannella C, Fernandez FR, Santoro F et al. Structure-Guided Design of Temporin-Derived Peptides Reveals Potent Dual-Mechanism Inhibitors of SARS-CoV-2. J Med Chem. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41622708/

[4] Hannula L, Kuivanen S, Lasham J et al. Nanobody engineering for SARS-CoV-2 neutralization and detection. Microbiol Spectr. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/38363137/

[5] Qin Q, Jiang X, Huo L et al. Computational design and engineering of self-assembling multivalent microproteins with therapeutic potential against SARS-CoV-2. J Nanobiotechnology. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/38341574/

[6] Jiang X, Qin Q, Zhu H et al. Structure-guided design of a trivalent nanobody cluster targeting SARS-CoV-2 spike protein. Int J Biol Macromol. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/38000614/

[7] Alwethaynani MS, Hassan RN, Hussain R et al. Structure-guided optimization of N-(5-((2-oxopropyl) thio)-1,3,4-thiadiazol-2-yl) propionamide frameworks as potent SARS-CoV-2 M(pro) inhibitor for treating coronavirus disease. Bioorg Chem. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/42314596/

[8] Nguyen HN, Ranasinghe PS, Ilesinghe IKRS et al. Structure-guided design of broad-spectrum inhibitors of coronaviral proteases embodying a 1,3,2-oxazaphospholidin-3-one scaffold as a versatile design element. Eur J Med Chem. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/42208369/

[9] van der Straat R, Oerlemans R, Cong Y et al. Ugi-Tetrazole-Derived α‑Aminomethyl Scaffolds Reveal Unexpected Binding Modes in SARS-CoV‑2 3CLpro. ACS Med Chem Lett. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41982734/

[10] Rahman MS, Han MJ, Kim SW et al. Structure-Guided Development of Bivalent Aptamers Blocking SARS-CoV-2 Infection. Molecules. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/37375202/

[11] Wang J, Xu Y, Yang Y et al. Structural Basis and Inhibitor Development of SARS-CoV-2 Papain-like Protease. Molecules. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41683451/

[12] Sharafi M, Teh WP, Green J et al. Structure-Guided Design of Potent and Selective Covalent Inhibitors Targeting the SARS-CoV-2 Papain-like Protease. J Med Chem. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41557701/

[13] Okabe A, Carney DW, Tawada M et al. Discovery of Highly Potent Noncovalent Inhibitors of SARS-CoV-2 Main Protease through Computer-Aided Drug Design. J Med Chem. 2025. URL: https://pubmed.ncbi.nlm.nih.gov/41076627/

[14] Bhagat A, Kelam LM, Samanta N et al. Structure-guided design and triplicate molecular dynamics evaluation of mutant peptide inhibitors targeting SARS-CoV-2 main protease (Mpro). In Silico Pharmacol. 2025. URL: https://pubmed.ncbi.nlm.nih.gov/40917523/

[15] El Bakri Y, Ahmad B, Saravanan K et al. Insight into crystal structures and identification of potential styrylthieno[2,3-b]pyridine-2-carboxamide derivatives against COVID-19 Mpro through structure-guided modeling and simulation approach. J Biomol Struct Dyn. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/37318002/

[16] Jha P, Saluja D, Chopra M. Structure-guided pharmacophore based virtual screening, docking, and molecular dynamics to discover repurposed drugs as novel inhibitors against endoribonuclease Nsp15 of SARS-CoV-2. J Biomol Struct Dyn. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/35652904/

[17] Rosas-Lemus M, Athe S, Minasov G et al. S‑Adenosylhomocysteine Analogs Selectively Suppress Pan-Coronavirus Replication by Inhibition of nsp14 Methyltransferase. ACS Med Chem Lett. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41568342/

[18] Mik V, Benz LS, Voller J et al. A patent review of Mpro protease inhibitors for the treatment of COVID-19 infections (2020 - present). Expert Opin Ther Pat. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41348159/

[19] Krishnathas R, Mineev KS, Fourkiotis NK et al. Structure-Based Rational Design of a Selective Hydrolase Inhibitor of the Severe Acute Respiratory Syndrome Coronavirus-2 Nsp3 Macrodomain. Chembiochem. 2025. URL: https://pubmed.ncbi.nlm.nih.gov/41176653/

[20] Stewart J, Jia R, Ali MA et al. Structure-Guided Temporin L Analogs Development to Inhibit the Main Protease of SARS-CoV‑2. ACS Med Chem Lett. 2025. URL: https://pubmed.ncbi.nlm.nih.gov/41089481/

[21] Shehzadi K, Kalsoom I, Yu MJ et al. Design and in-silico evaluation of PNA-based novel pronucleotide analogues targeting RNA-dependent RNA polymerase to combat COVID-19. J Biomol Struct Dyn. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/39937582/

[22] Kuhn AJ, Outlaw VK, Marcink TC et al. Enhancing the solubility of SARS-CoV-2 inhibitors to increase future prospects for clinical development. J Virol. 2025. URL: https://pubmed.ncbi.nlm.nih.gov/39902960/

[23] Yin DH, Xin J, Chen S et al. Structure-guided design and photochemical synthesis of new carbamo(dithioperoxo)thioates with improved potencies to SARS-CoV-2 3CL(pro). Bioorg Med Chem. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/39442491/

[24] Dampalla CS, Kim Y, Zabiegala A et al. Structure-Guided Design of Potent Coronavirus Inhibitors with a 2-Pyrrolidone Scaffold: Biochemical, Crystallographic, and Virological Studies. J Med Chem. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/38953866/

[25] Papini C, Ullah I, Ranjan AP et al. Proof-of-concept studies with a computationally designed M(pro) inhibitor as a synergistic combination regimen alternative to Paxlovid. Proc Natl Acad Sci U S A. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/38621119/

[26] Westberg M, Su Y, Zou X et al. An orally bioavailable SARS-CoV-2 main protease inhibitor exhibits improved affinity and reduced sensitivity to mutations. Sci Transl Med. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/38478629/

[27] Shehzadi K, Yu M, Liang J. De Novo Potent Peptide Nucleic Acid Antisense Oligomer Inhibitors Targeting SARS-CoV-2 RNA-Dependent RNA Polymerase via Structure-Guided Drug Design. Int J Mol Sci. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/38139312/

[28] Inniss NL, Kozic J, Li F et al. Discovery of a Druggable, Cryptic Pocket in SARS-CoV-2 nsp16 Using Allosteric Inhibitors. ACS Infect Dis. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/37728236/

[29] Kang KM, Jang Y, Lee SS et al. Discovery of antiviral SARS-CoV-2 main protease inhibitors by structure-guided hit-to-lead optimization of carmofur. Eur J Med Chem. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/37633203/

[30] Hausdorff M, Delpal A, Barelier S et al. Structure-guided optimization of adenosine mimetics as selective and potent inhibitors of coronavirus nsp14 N7-methyltransferases. Eur J Med Chem. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/37192550/

[31] Dampalla CS, Miller MJ, Kim Y et al. Structure-guided design of direct-acting antivirals that exploit the gem-dimethyl effect and potently inhibit 3CL proteases of severe acute respiratory syndrome Coronavirus-2 (SARS-CoV-2) and middle east respiratory syndrome coronavirus (MERS-CoV). Eur J Med Chem. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/37080108/

[32] Bobileva O, Bobrovs R, Sirma EE et al. 3-(Adenosylthio)benzoic Acid Derivatives as SARS-CoV-2 Nsp14 Methyltransferase Inhibitors. Molecules. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/36677825/

[33] Dampalla CS, Nguyen HN, Rathnayake AD et al. Broad-Spectrum Cyclopropane-Based Inhibitors of Coronavirus 3C-like Proteases: Biochemical, Structural, and Virological Studies. ACS Pharmacol Transl Sci. 2023. URL: https://pubmed.ncbi.nlm.nih.gov/36654747/

[34] Dampalla CS, Rathnayake AD, Galasiti Kankanamalage AC et al. Structure-Guided Design of Potent Spirocyclic Inhibitors of Severe Acute Respiratory Syndrome Coronavirus-2 3C-like Protease. J Med Chem. 2022. URL: https://pubmed.ncbi.nlm.nih.gov/35638577/

[35] Nencka R, Silhan J, Klima M et al. Coronaviral RNA-methyltransferases: function, structure and inhibition. Nucleic Acids Res. 2022. URL: https://pubmed.ncbi.nlm.nih.gov/35018474/ *** Disclaimer: This article is for educational and informational purposes only. It is not intended to substitute for professional veterinary advice, diagnosis, treatment, or regulatory guidance. Always consult a licensed veterinarian or qualified specialist regarding animal health, disease diagnosis, and therapeutic decisions.