key: cord-0726822-bv90ywp6
authors: Priya, Prerna; Basit, Abdul; Bandyopadhyay, Pradipta
title: A strategy to optimize the peptide-based inhibitors against different mutants of the spike protein of SARS-CoV-2
date: 2022-02-28
journal: bioRxiv
DOI: 10.1101/2022.02.27.482153
sha: a0a633a94feea1f9b2a318465b039be727e9db04
doc_id: 726822
cord_uid: bv90ywp6

SARS-CoV-2 virus has caused high-priority health concerns at a global level. Vaccines have stalled the proliferation of viruses to some extent. Yet, the emergence of newer, potentially more infectious, and dangerous mutants such as delta and omicron are among the major challenges in finding a more permanent solution for this pandemic. The effectiveness of antivirals Molnupiravir and Paxlovid, authorized for emergency use by the FDA, are yet to be assessed at larger populations. Patients with a high risk of disease progression or hospitalization have received treatment with a combination of antibodies (antibody-cocktail). Most of the mutations leading to the new lineage of SARS-CoV-2 are found in the spike protein of this virus that plays a key role in facilitating host entry. The current study has investigated how to modify a promising peptide-based inhibitor of spike protein, LCB3, against common mutations in the target protein so that it retains its efficacy against the spike protein. LCB3 being a prototype for protein-based inhibitors is an ideal testing system to learn about protein-based inhibitors. Two common mutations N501Y and K417N are considered in this work. Using a structure-based approach that considers free energy decomposition of residues, distance, and the interactions between amino acids, we propose the substitutions of amino acid residues of LCB3 inhibitors. Our binding free energy calculations suggest a possible improvement in the binding affinity of existing inhibitor LCB3 to the mutant forms of the S-protein using simple substitutions at specific positions of the inhibitor. This approach, being general, can be used in different inhibitors and other mutations and help in fighting against SARS-CoV-2.

leads to the viral spread [39] [40] . GRL0617 is a promising inhibitor against PLpro 41 . Emergency use authorization has been given to several vaccines, monoclonal antibodies as well as antiviral drugs developed by various pharmaceutical companies to save lives during the pandemic [42] [43] [44] [45] . As a result, more than 10 billion doses of vaccine have been already administered to curb the progress of COVID-19 (https://covid19.who.int/). Remdesivir, an antiviral drug, was used widely but did not show significant clinical benefit against COVID-19 46 . Monoclonal antibodies 'sotrovimab' as well as the antibody cocktaila combination of 'casirivimab' and 'imdevimab' have shown clinical benefits and reduced the hospitalization rate in patients with (https://www.covid19treatmentguidelines.nih.gov/) 45, [47] [48] . However, none of these has offered a real cure against this disease yet. People are also getting infected, post-vaccination, with new variants though showing milder infection 49 . Recently, FDA authorized two COVID-19 antiviral pills: 'Molnupiravir' (Merck, USA) and, 'Paxlovid' (Pfizer, USA) for emergency use in patients who are at high risk [50] [51] [52] . Though these pills worked well in clinical trials, the real-world efficacies of these pills are yet to be assessed.

One of the major challenges of treating COVID-19 is the appearance of different mutant strains of the virus. The mutation frequency of SARS-CoV-2 to form newer strains such as alpha, beta, a more aggressive delta, and the most recent omicron, is a major challenge in developing a treatment for COVID [53] [54] [55] [56] . It also includes the question, of whether the treatments developed for SARS-CoV-2 will work on its existing and upcoming variants [57] [58] [59] . The majority of the mutations reported are in the spike (S) glycoprotein, which is responsible for the entry of the virus inside the human host via human receptor protein Angiotensin Converting Enzyme-2 (ACE-2) [60] [61] [62] . In this work, we want to check if existing inhibitors can be modified in such a way that these retain their efficacy for the mutant forms. For a proof-of-principle study, we have taken a designed mini-protein inhibitor, LCB3, a prototype for protein-based inhibitors including antibodies 12 (Figure 1 ). LCB3 is a 64 amino-acid-long, stable, potent miniprotein inhibitor that competes with ACE-2 and binds tighter with the S-protein 12 . It has the potential to be used as a therapeutic and may open options for direct delivery to the nasal passage or other parts of the respiratory system. However, several studies have shown mutations induced alteration in the binding affinity of S-protein to ACE-2 receptor [63] [64] [65] [66] . SN501Y (Spike protein of SARS-CoV-2 with N501Y mutation) binds to ACE-2 receptor with 7fold higher affinity than the WT whereas SK417N (Spike protein of SARS-CoV-2 with K417N mutation) binds with 4-fold lower affinity 67 . This reflects the possibility of alteration in the binding affinity of the LCB3 inhibitors against the mutated S-protein and needs further investigation. Two common mutations, N501Y and K417N are present at the binding interface of the protein ( Figure   1 ) were tested. We have investigated the effect of these two mutations on the binding affinity between LCB3 and S-protein and how to improve the binding affinity by modifying the inhibitors.

N501Y mutation is found in various lineages including B.1.1.7 (alpha), B.1.351 (beta), and P.1 (gamma) variants first detected in the U.K., South Africa and Brazil respectively 68 69 . Both of these mutations (N510Y and K417N) were also reported in the newly detected variant of concern, lineage B.1.1.529 (omicron) 70 . These mutations have immune evasion properties [71] [72] [73] .

To investigate the potential change in the binding affinity, associated with the mutations N501Y and K417N of S-protein with LCB3 and to improve the LCB3 inhibitor to make it effective against WT as well as mutants, we have used binding free energy calculation with the well-known Molecular Mechanics Poisson-Boltzmann Surface Area (MM-PBSA) method. We have decomposed the results of the binding affinity per residue of the inhibitor to see which residues contribute most to the binding. Then residues of LCB3 that are responsible for less binding affinity are changed to other residues based on the type of interaction and binding site geometry. This simple procedure increases the binding affinity of modified LCB3 with the two mutant forms of the proteins. This procedure is general and can be used to optimize other inhibitors against other mutations as well. 

: ff14SB force field of AMBER 16 package was used to generate the parameters of the proteins [74] [75] . All systems were solvated using TIP3P water molecules in a rectangular box using a minimum 10 Å distance between the edges of the box to the surface of the protein [76] [77] . Special care was taken to preserve the disulfide bonds present between the cysteine residues present in PDB. Four pairs of Cysteines (C336-C361, C379-C432, C480-C488, C391-C525) of the S protein were involved in the disulfide bonding. Counter ions were added to neutralize the system and to maintain the 0.15M KCl salt concentration. 10000 steps of steepest descent followed by 10000 conjugant gradient minimization was used to remove the bad contacts in the solvated system [78] [79] . The systems were heated slowly up to 298 K for 50 ps followed by 50 ps density equilibration with the restraint weight of 2 kcal/mol/Å 2 followed by and 500 ps equilibration run. After equilibration 10 independent simulations of all the complex, each of 10 ns was performed at NPT (by maintaining 300K temperature and 1 atmospheric pressure) to increase the sampling space (Table S1 ). For entropy calculations, separate long simulations (1 μs) of complex, receptor and ligands were performed (Table S2 ). Hydrogen bonds were constrained using a shake algorithm with 2 fs time integration 80 . The temperature was regulated by Langevin Dynamics with 2 ps of relaxation time 81 . 1 bar pressure was controlled by Berendesen's barostat and periodic boundary conditions were applied for all the systems 82 . The particle mesh Ewald summation method was used for the calculation of the long-range electrostatic 83 . CPPTRAJ was used for further analysis of the trajectories obtained from simulation 84 . 

Here, ΔGcomplex, ΔGreceptor, ΔGligand are the free energy of solvation for complex receptor, and ligand respectively. < > denotes the ensemble average.

ΔGbinding is calculated by equation 2:

ΔGgas is the change in the gaseous state. It is calculated by equation 5. 86 .

The nonpolar component is calculated using equation 6.

The value of γ and β is 0.0378 kcal/mol-Å 2 and -0.569 kcal/mol respectively as implemented in the AMBER package. The solvent-accessible surface area (SASA) is used to distinguish the exposed and buried area of the protein. For that, a probe sphere with vdW radii of ~1.4 Å is rolled along the surface and if the probe can cross the area it is considered as surface accessible else the region is considered as buried where the solvent cannot enter.

The 0.15M ionic concentration was used for MM-PBSA calculations. The dielectric constant of the solute is highly dependent on the characteristics of the investigated system [87] [88] . In this study, various dielectric constants of solute were tested and the selection was done based on the similarity with the available data (details given in supporting information, Table S3 ). The dielectric constant of solute was kept 8. The total binding energy was averaged using 1000 frames obtained from 10 independent trajectories, each of 10 ns, for all the systems.

The entropy of the complex, receptor and ligand were estimated using Quasi Harmonic (QH) approximation method using CPPTRAJ. A sufficient phase space sampling is required for the reliable estimation of entropic contributions 89 . For that, an overall 11 μs simulation was performed (Table S2) . To estimate the effects of mutation on the flexibility of the system, entropy change (TΔS) was calculated using 5 lakhs frames from the separate 1 μs simulation, of complex, receptor and ligand.

The binding free energies of SWT (wild-type S-protein), SN501Y (asparagine at 501 position of Sprotein mutated with tyrosine) and SK417N (lysine at 417 position of S-protein mutated with asparagine) with LCB3 were determined using the MM-PBSA calculations and the values were compared. The binding affinity of SK417N and SN501Y mutants were reduced as compared with SWT by ~ +8 kcal/mol (Table 1) .

The loss of +7.1 kcal/mol binding affinity of N501Y mutant of spike protein from WT to the mutant with LCB3 is reported by Williams et al. reflecting the correctness of our calculation protocol. 90 The SN501Y mutation results in a loss of binding due to nonpolar solvation interaction (+6.44 kcal/mol). There is also a loss of +2.95 kcal/mol in polar contributions accompanied by the gain in the van der Waals (vdW) component by -1.23 kcal/mol (Table 1) . The lower binding affinity of SK417N mutant is mainly due to the loss in polar interactions (+6.01 kcal/mol) and non-polar solvation (~+5.01 kcal/mol) with the slight gain of -2.80 kcal/mol of vdW interactions. It reflects that when the positively charged residue lysine was mutated with the uncharged polar asparagine residue, the polar interactions reduced.

The reduced binding affinity of LCB3 with SK417N and SN501Y mutants provided the scope to improve the LCB3 inhibitor further in such a way so that it can work potently and inhibit the WT as well as mutants. For that, residue-wise free energy decomposition was performed for the identification of important residues 91 . Free energy contributions of each residue were decomposed into electrostatic, vdW, and polar solvation and the amino residues of LCB3 near the mutated residue of S-protein were analyzed ( Table 2 ). D3 (Table 2) . However, the mutation (SN501Y or SK417N) leads to the overall loss of ~+8 kcal/mol (Table 1 ). The loss in the energy contributions of E4 (+1.3 kcal/mol in SN501Y and +0.5 kcal/mol in SK417N), T10 (+0.2 kcal/mol in SN501Y and +0.7 kcal/mol in SK417N) and D11 (+0.9 kcal/mol in SN501Y and +0.5 kcal/mol in SK417N) were observed in both SN501Y and SK417N (Table 2 ). It is likely that the interactions between these residues of LCB3 and neighboring residues of the S-protein have the maximum contribution to the loss in binding affinity. Detailed view of the binding interface: the interacting residues are shown in the stick representation. The distance among the oppositely charged atoms which may interact is given in Å.

Apart from the free energy contributions, it is also essential to understand the mutations induced conformational changes and interaction patterns to improve the binding affinity. In SWT the N1, D3, and E4 of LCB3 form polar interaction with Q498 and N501 of S-protein (Figure 2 and 3a) .

The replacement of polar asparagine by hydrophobic tyrosine (N501Y) disrupts the electrostatic contribution (Figure 3b ) of asparagine which was reflected in Table 1 . In the structure of S-protein with ACE2 receptor, the polar residues are surrounded by two tyrosine residues Y41 of ACE2 and Y505 of S-protein (Figure 3c ) 92 which was lacking with LCB3 inhibitor. In SK417N, the replacement of a long side-chain residue lysine with asparagine weakens the interaction between SK417N and LCB3 inhibitor (Figure 2, 4a, and 4b) . This conformational change leads to the overall loss of +8.22 kcal/mol in the binding affinity of SK417N with LCB3 inhibitor (Table 1 ). Figure 5b ). To overcome the loss in the binding affinity of SK417N, two residues of LCB3: T10, and D11 were targeted for substitutions, based on the loss in the free energy contributions of the residues, and the distance from the mutated K417 (Table 2, Figure 2 and 4) . Substitutions of the amino acid T10 of LCB3 by D, E and Q were tested with the possibility to improve polar interaction with N417 but none of these have enhanced the affinity (Table 3 ). D11 is closer to K417 than T10 (Figure 2 ), forms direct polar interaction with K417 (Figure 2b) , and there is a loss of +0.5 kcal/mol in the free energy contributions of D11 in SK417N as compared to SWT (Table 2) . Therefore, D11 was replaced with a bulky and long side-chain amino acid that can interact with N417. In this regard, two substitutions, D11R and D11H were tested. The binding affinity of the LCB3D11R with K417N

is -17.35 kcal/mol, ~+6.5 kcal/mol lesser affinity than the SWT-LCB3 (Table 3) . However, LCB3D11H showed a higher binding affinity (-23.38 kcal/mol) with SK417N (Table 3) . Here, two different protonation states of histidine, at ND1 (HID) and HE2 (HIE) were tested. Histidine with protonation at HE2 (HIE) has shown higher binding affinity as compared to HID. 'H' refers to the 'HIE' protonation state in this paper. From the structure, it was found that the imidazole ring moves towards the binding interface and comes closer to N417 (Figure 6 a) , and strengthens the binding by improving the nonpolar solvation component ( Table 2 ). The binding affinity of LCB3D11H with SWT is -22.33 kcal/mol, slightly less than the binding affinity of LCB3. In SWT-LCB3D11H, K417

of S-protein repels the H11 slightly away from the binding interface ( Figure 6b ). In the binding free energy calculations described so far, the entropy of solute is not considered, a common practice in MM-PBSA calculations. However, for the present problem, the calculation of solute entropy is important as changes in amino acids in either S-protein or LCB3 can change the flexibility of the system leading to a change in entropy. As the entropy calculations are computationally intensive, we have performed it for some important systems as mentioned in Table 4 . The change in the entropy was calculated over separate 1μs trajectories complex, receptor and ligand. The entropy essentially converges for each system in 1μs MD simulation (Figure 7a to c). The negative of the change in entropy multiplied by temperature, T (the last term of equation (3), -TΔS) of LCB3 (+55.5 kcal/mol) and its two variants LCB3D3Y (+56.0 kcal/mol) and LCB3D11H (+56.2 kcal/mol) with SWT were almost similar with no significant differences between them ( Figure 7d , Table S4 ). However, the significant differences in -TΔS were observed for SN501Y-LCB3D3Y (+49.4 kcal/mol) and SK417N-LCB3D11H (+24.4 kcal/mol). In both these cases, the entropy change on binding was significantly reduced. This reduction in the change of entropy will also help in improving the binding affinity of these variants of LCB3 with mutated S-protein (equation 3). It reflects that the binding affinity of SWT with the proposed variants of LCB3 is mainly enthalpy driven; however, in the mutated S-protein (SN501Y and SK417N) entropic changes played significant contributions in improving the binding affinity (Table 4 ). To summarize, our calculation and analysis show that a judicious combination of free energy decomposition and geometric consideration can suggest ways to improve the inhibitor against specific mutations. Detailed binding affinity calculation that includes solute entropy shows that even a single amino acid change in LCB3 can make it a potent inhibitor to the S-protein and its mutants. In particular, LCB3D3Y and LCB3D11H were proposed as the variants of LCB3 that may inhibit the WT as well as mutants of S-protein. It is to be noted that LCB3 is taken as a prototype for more complex systems such as antibodies. Moreover, this methodology is general so that it can be applied to any inhibitor against different targets.

Our calculation methodology does have some caveats. First, as we wanted to develop a fast methodology, only the receptor-binding domain of the S-protein and the LCB inhibitor were considered. The ACE-2 protein is not considered in the calculation. Our choice of using a simpler system is similar to the work done by Williams et al 90 .

The fight against SARS-CoV-2 requires further development of vaccines and medicines.

One of the main targets to develop a drug against this virus is the spike protein of the virus.

Although there are promising drug candidates against the virus, the major issue is how to design/modify inhibitors against the various mutant strains of the virus. In this work, as a proofof-principle study, we have considered a mini-protein inhibitor, LCB3 that binds to the spike protein of SARS-CoV-2. We have devised a computational protocol to modify LCB3 against the two most common mutations in the spike protein, namely, N501Y and K417N. Our computational methodology includes a free energy decomposition procedure in conjunction with a detailed analysis of the binding site. Our proposed modified LCB3 is shown to bind to the two mutant forms of the spike protein potently. The binding enthalpy showed that the single residue mutations LCB3D3Y work well against SWT as well as SN501Y and SK417N. Another modification LCB3D11H binds efficiently with SK417N and SWT but not with SN501Y. Further, the entropic contributions also favor the binding of LCB3D3Y with SN501Y, and LCB3D11H with SK417N respectively. The study found that LCB3D3Y and LCB3D11H were bound with a higher affinity with the WT as well as mutated Sprotein. This strategy can be useful to redesign the peptide-based inhibitor against the target protein that undergoes frequent mutation.

Coronavirus biology and replication: implications for SARS-CoV-2

Characteristics of SARS-CoV-2 and COVID-19

Transmission of SARS

CoV-2: A Review of Viral, Host, and Environmental Factors

The architecture of the SARS-CoV-2 RNA genome inside virion

The coding capacity of SARS-CoV-2

Potential drug targets of SARS-CoV-2: From genomics to therapeutics

Polymerase Are Effective Targets of Therapeutic Intervention for Treatment of COVID-19 Caused by SARS-CoV-2 Variants (B.1.1.7 and B.1.351). Microbiol Spectr

Quinolines-Based SARS

CoV-2 3CLpro and RdRp Inhibitors and Spike-RBD-ACE2 Inhibitor for Drug-Repurposing Against COVID-19: An in silico Analysis

Identification of a New Potential SARS-COV-2 RNA-Dependent RNA Polymerase Inhibitor via Combining Fragment-Based Drug Design, Docking, Molecular Dynamics, and MM-PBSA Calculations

Recent advances in developing small-molecule inhibitors against SARS-CoV-2

De Novo Discovery of High-Affinity Peptide Binders for the SARS-CoV-2 Spike Protein

De novo design of picomolar SARS-CoV-2 miniprotein inhibitors

Salvianolic acid C potently inhibits SARS-CoV-2 infection by blocking the formation of sixhelix bundle core of spike protein

The anti-influenza virus drug, arbidol is an efficient inhibitor of SARS-CoV-2 in vitro

Efficacy and safety of arbidol (umifenovir) in patients with COVID-19: A systematic review and meta-analysis

Umifenovir in hospitalized moderate to severe COVID-19 patients: A randomized clinical trial

Small-Molecule Inhibitors of the Coronavirus Spike: ACE2 Protein-Protein Interaction as Blockers of Viral Attachment and Entry for SARS-CoV-2

Identifying SARS-CoV-2 Entry Inhibitors through Drug Repurposing Screens of SARS-S and MERS-S Pseudotyped Particles

A novel class of TMPRSS2 inhibitors potently block SARS-CoV-2 and MERS-CoV viral entry and protect human epithelial lung cells

Comparative analysis of antiviral efficacy of FDAapproved drugs against SARS-CoV-2 in human lung cells

Camostat mesilate therapy for COVID-19

Camostat mesylate inhibits SARS-CoV-2 activation by TMPRSS2-related proteases and its metabolite GBPA exerts antiviral activity

Efficacy of the TMPRSS2 inhibitor camostat mesilate in patients hospitalized with Covid-19-a double-blind randomized controlled trial

Nafamostat Mesylate Blocks Activation of SARS-CoV-2: New Treatment Option for COVID

Nafamostat in hospitalized patients with moderate to severe COVID-19 pneumonia: a randomised Phase II clinical trial

Marine Natural Products for Discovering TMPRSS2 Inhibitors. Front Chem 2021

Structure and drug binding of the SARS-CoV-2 envelope protein transmembrane domain in lipid bilayers

Coronavirus envelope protein: current knowledge

Inhibition of SARS CoV Envelope Protein by Flavonoids and Classical Viroporin Inhibitors

Structural Basis of Potential Inhibitors Targeting SARS-CoV-2 Main Protease

Identification of 14 Known Drugs as Inhibitors of the Main Protease of SARS-CoV-2

Structure of M(pro) from SARS-CoV-2 and discovery of its inhibitors

The main protease and RNA-dependent RNA polymerase are two prime targets for SARS-CoV-2

Structure of replicating SARS-CoV-2 polymerase

Mechanism of SARS-CoV-2 polymerase stalling by remdesivir

Structure of the SARS-CoV-2 RNA-dependent RNA polymerase in the presence of favipiravir-RTP

RNA-dependent RNA polymerase: Structure, mechanism, and drug discovery for COVID-19

Papain-like protease regulates SARS-CoV-2 viral spread and innate immunity

Structure of papain-like protease from SARS-CoV-2 and its complexes with non-covalent inhibitors

The complex structure of GRL0617 and SARS-CoV-2 PLpro reveals a hot spot for antiviral drug discovery

Emergency Use Authorizations of COVID-19-Related Medical Products

Emergency Use Authorization of Covid Vaccines -Safety and Efficacy Follow-up Considerations

Remdesivir: A Review of Its Discovery and Development Leading to Emergency Use Authorization for Treatment of COVID-19

Casirivimab-Imdevimab treatment is associated with reduced rates of hospitalization among highrisk patients with mild to moderate coronavirus disease-19

Repurposed Antiviral Drugs for Covid-19 -Interim WHO Solidarity Trial Results

Monoclonal Antibodies for COVID-19

Early Treatment for Covid-19 with SARS-CoV-2 Neutralizing Antibody Sotrovimab

Risk factors and disease profile of post-vaccination SARS-CoV-2 infection in UK users of the COVID Symptom Study app: a prospective, community-based

Molnupiravir -A Step toward Orally Bioavailable Therapies for Covid-19

Molnupiravir, an Oral Antiviral Treatment for COVID-19

Covid-19: Pfizer's paxlovid is 89% effective in patients at risk of serious illness, company reports

SARS-CoV-2 variants, spike mutations and immune escape

Omicron SARS-CoV-2 variant: a new chapter in the COVID-19 pandemic

The emerging plasticity of SARS-CoV-2

Not all COVID-19 pandemic waves are alike

Will Mutations in the Spike Protein of SARS-CoV-2 Lead to the Failure of COVID-19 Vaccines?

The Impact on Infectivity and Neutralization Efficiency of SARS-CoV-2 Lineage B.1.351 Pseudovirus

Anti-SARS-CoV-2 Vaccines and Monoclonal Antibodies Facing Viral Variants

Human SARS CoV-2 spike protein mutations

Large-scale analysis of SARS-CoV-2 spike-glycoprotein mutants demonstrates the need for continuous screening of virus isolates

Emerging SARS-CoV-2 variants of concern and potential intervention approaches

N501Y mutation of spike protein in SARS-CoV-2 strengthens its binding to receptor ACE2

Higher infectivity of the SARS-CoV-2 new variants is associated with K417N/T, E484K, and N501Y mutants: An insight from structural data

SARS-CoV-2 D614G spike mutation increases entry efficiency with enhanced ACE2-binding affinity

The new SARS-CoV-2 strain shows a stronger binding affinity to ACE2 due to N501Y mutant

Experimental Evidence for Enhanced Receptor Binding by Rapidly Spreading SARS-CoV-2 Variants

Addendum: A dynamic nomenclature proposal for SARS-CoV-2 lineages to assist genomic epidemiology

What's important to know about SARS-CoV-2 variants of concern?

SARS-CoV-2 Omicron strain exhibits potent capabilities for immune evasion and viral entrance

Neutralising antibody escape of SARS-CoV-2 spike protein: Risk assessment for antibody-based Covid-19 therapeutics and vaccines

N501Y and K417N Mutations in the Spike Protein of SARS-CoV-2 Alter the Interactions with Both hACE2 and Human-Derived Antibody: A Free Energy of Perturbation Retrospective Study

Increased immune escape of the new SARS-CoV-2 variant of concern Omicron

Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB

Comparison of simple potential functions for simulating liquid water

The Amber biomolecular simulation programs

The Origin of the Method of Steepest Descent

On the iterative methods for solving linear systems of equations

Rattle: A "velocity" version of the shake algorithm for molecular dynamics calculations

Langevin stabilization of molecular dynamics

Molecular Dynamics with Coupling to an External Bath

An N•log(N) Method for Ewald Sums in Large Systems

Software for Processing and Analysis of Molecular Dynamics Trajectory Data

py: An Efficient Program for End-State Free Energy Calculations

Implicit Nonpolar Solvent Models

Assessing the performance of the MM/PBSA and MM/GBSA methods. 10. Impacts of enhanced sampling and variable dielectric model on protein-protein Interactions

Assessing the performance of the MM/PBSA and MM/GBSA methods. 1. The accuracy of binding free energy calculations based on molecular dynamics simulations

Method for estimating the configurational entropy of macromolecules

Fast Prediction of Binding Affinities of the SARS-CoV-2

Spike Protein Mutant N501Y (UK Variant) with ACE2 and Miniprotein Drug Candidates

Insights into protein-protein binding by binding free energy calculation and free energy decomposition for the Ras-Raf and Ras-RalGDS complexes

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor