key: cord-0986777-5obbi9qh
authors: Dutta, Kunal; Elmezayen, Ammar D.; Al-Obaidi, Anas; Zhu, Wei; Morozova, Olga V.; Shityakov, Sergey; Khalifa, Ibrahim
title: Seq12, Seq12m, and Seq13m, peptide analogues of the spike glycoprotein shows antiviral properties against SARS-CoV-2: An in silico study through molecular docking, molecular dynamics simulation, and MM-PB/GBSA calculations
date: 2021-07-16
journal: J Mol Struct
DOI: 10.1016/j.molstruc.2021.131113
sha: 5a9e61362466fbd9982b91900448787ae829443d
doc_id: 986777
cord_uid: 5obbi9qh

At the very beginning of the new decade, the COVID-19 pandemic has badly hit modern human societies. SARS-CoV-2, the causative agent of COVID-19 acquiring mutations and circulating as new variants. Herein, we have found three new antiviral peptides (AVPs) against the SARS-CoV-2. These AVPs are analogous to the spike glycoprotein of the SARS-CoV-2. Antiviral peptides, i.e., Seq12, Seq12m, and Seq13m, can block the receptor-binding domain (RBD) of the SARS-CoV-2, which is necessary for communicating with the angiotensin-converting enzyme 2 (ACE2). Also, these AVPs sustain their antiviral properties, even after the insertion of 25 mutations in the RBD (Rosetta and FoldX based). Further, Seq12 and Seq12m showed negligible cytotoxicity. Besides, the binding free energies calculated using MM-PB/GBSA method are also in agreement with the molecular docking studies. The molecular interactions between AVPs and the viral membrane protein (M) also showed a favorable interaction suggesting it could inhibit the viral re-packaging process. In conclusion, this study suggests Seq12, Seq12m, and Seq13m could be helpful to fight against SARS-CoV-2. These AVPs could also aid virus diagnostic tools and nasal spray against SARS-CoV-2 in the future.

At present, the entire World is facing challenges to handle the COVID-19 pandemic [1] .

SARS-CoV-2 is the causative agent of COVID-19, and it is new to the scientific community.

Therefore, in-depth monitoring of all aspects of the COVID-19 is ongoing. For example, signs and symptoms [2] , mode of transmission [3] , WHO-solidarity trials [4] , contact tracing by mobile apps such as "Arogya Setu" by India [5] , CRISPR based rapid diagnostic of SARS-CoV-2 [6] , and also monitoring daily cases by crowd-sourcing (https://www.covid19india.org/). Besides, recent reports suggest that repurposing known antiviral drugs [7] , drugs [8] , different phytochemicals [9] against COVID-19 could be fruitful. However, none of them has reached a final definitive clinical treatment for COVID- 19 . Indeed to address the urgent need for a safe and efficacious vaccine against the COVID-19 several vibrant initiatives have been started as never before. For example, vaccine manufacturing front-runner come-up with mRNA vaccines [10] , viral vector vaccine [11] , classical attenuated vaccine etc. [12] . However, reports showed COVID-19 vaccines would not be a silver bullet for the immunization of a community.

Furthermore, alternatives to traditional therapeutics would be necessary for the long run as before [13, 14] . Moreover, SARS-CoV-2 acquires new mutations in its genome in a concise time frame [15] , for instance, B.1.1.7, B.1.351, P.1, B.1.427, and B.1.429, etc. [16] . These SARS-CoV-2 variants are one of the main concerns for all anti-COVID-19 efforts [17] .

SARS-CoV-2 encodes a spike glycoprotein that assembled as a trimer for function. The notorious entry of SARS-CoV-2 inside a human host is primarily mediated via a proteinprotein interaction between angiotensin-converting enzyme 2 (ACE2) and the receptorbinding domain (RBD) of the SARS-CoV-2 [18] . Therefore, RBD holds importance as a potential pharmaceutical target. Herein, we made an effort to find out new antiviral peptides against SARS-CoV-2. In such an effort, we utilized Machine Learning and Supported Vector Machine for antiviral peptide predictions. After that, we opt for residue-resistant molecular docking, molecular dynamics simulations, and MM-PB/GBSA analysis to characterize the antiviral peptides and the molecular interactions between AVP-RBD complexes. Hydrogen bonds (H-bonds) are weak non-covalent interactions with a binding strength of one-tenth than a typical covalent bond [76] . However, H-bonds are essential for proteins and nucleic acids function [76, 77] . And this study is no exception, as many H-bonds were formed during the molecular dynamics trajectory between the AVP-RBD complex and water. Best of our knowledge, this is the first report of antiviral peptides against the SARS-CoV-2 derived from its RBD of the spike glycoprotein.

Nucleotide sequences of the SARS-CoV-2 were obtained from NCBI viruses [19] . The complete genome sequence of the SARS-CoV-2 strain Wuhan-Hu-1 (GenBank Sequence Accession: MN908947) was used as the query sequence. The viral spike glycoprotein (S-Protein) and membrane protein (M-Protein) (QHD43416.1) were studied using PSI-BLAST.

Homology models of the M-and S-Protein were built using i-TASSER and SWISS-MODEL, respectively [20, 21] , followed by structure validation using PROCHECK [22] .

The aligned sequences of spike protein resulted from the PSI-BLAST were used to discover motifs using the MEME suite [23, 24] . The primary query included a set of 69 protein sequences, between 16 to 1196 amino acids in length (average length 214.3 amino acids), with the following settings: Background: A order-0 background generated from the supplied sequences. Discovery Mode: Classic: optimize the E-value of the motif information content. 

Identified motifs were checked and validated as antiviral peptide using a Machine Learning algorithm (MLA), and Supported Vector Machine (SVM), i.e., Meta-iAVP [25] , and AVPred [26] with four prediction model settings viz., AVP motif, sequence alignment, composition analysis, physio-chemical properties (threshold value of 50). The theoretical half-maximum inhibitory concentration (IC 50 ) value was calculated using AVP-IC 50 Pred [27] .

The predicted antiviral peptide sequences were used to build three-dimensional (3D) structures using the PEP-FOLD (v3.5) [28] . The structures obtained were then subject to energy minimization (GROMACS, v5.0) to remove unusual torsions and clashes from their structural geometry [29] . Properties of the predicted antiviral peptides were calculated using Innovagen peptide property calculator [30] , ProtParam [31] , AIPpred [53] AlgPred (Mapping of IgE epitopes and PID, MEME/MAST motif, Blast search on allergen representative peptides (ARPs)), [32] , ToxinPred (SVM (TrEMBL) + Motif based and SVM (Swiss-Prot) based), [33] , HemoPred, (SVM + Motif (HemoPI-2) based), IL4pred (Hybrid (SVM + motif) based ), [34] . AVP-IC 50 Pred (Hybrid Model Features a: Composition (mono-di) + Physico + Secondary structure + Surface accessibility and Features b: Binary (N8/C8) + Physico + Secondary structure + Surface accessibility) was used by selecting a Machine Learning Techniques: Supported Vector Machine (SVMlight) Random Forest (R package), IBk (Weka), KStar (Weka), [27] .

Crystal structure of the receptor-binding domain (6W41) of the SARS-CoV-2 was obtained from RCSB-PDB [35] . We manually select Chain C of the 6W41 as it is annotated as RBD for ACE2. The in silico multi-point mutant models were build using FireProt [36] .

HADDOCK-v2.2 was used for molecular docking between AVPs and receptor-binding domain (RBD) of the viral spike glycoprotein, viral membrane protein (M), RNA-dependent-RNA-polymerase (RdRp) [37] . Results were visualized using Discovery Studio Visualizer [38] .

The MD-simulation of the antiviral peptides, i.e., Seq12, Seq12m, and Seq13m, were carried out for 50 ns using GROMACS (v2019) [39] . In addition, 100 ns MD-simulation studies of the AVP-RBD complexes were performed using NAMD as described previously [8, 40] . In brief, input files were generated using the web-based server CHARMM-GUI (http://www.charmm-gui.org/). All systems were neutralized with KCl at 0.15 Molar, using the Monte-Carlo ion placing method. Systems were solvated using TIP3 water model, and

CHARMM36m force field was used to assign charges. The MD simulations were performed in two stages. At the first stage of the MD run, the energy minimization was performed for 20,000 steps by the steepest descent method followed by restrained 5 ns-equilibration at NVT ensemble. We did not provide any trajectory analysis for this stage as it is a restrained system and would be a biased and unreliable examination in our study. At the second stage, unrestrained 100 ns production at 310 K was performed under the NPT ensemble. RMSD, RMSF, Radius of gyration (Rg), H-bonds, and SASA were analyzed using VMD. The average values of multiple replicated MD simulations were used for analysis. A detailed method of the MD simulation is described in the supplementary section. were calculated with NMODE analysis tool in AMBER 16 package [42, 43] . Additionally, HawkDock was also used for MM-GBSA free energy calculations (per amino acid residues) [44] . The results were analyzed using GraphPad Prism 6 (San Diego, CA, USA). A detailed method is described in the supplementary section.

Antiviral peptides were checked for their immunogenic properties using the IEDB epitope analysis tool, which includes sequential B-cell epitopes [45] , T-cell (MHC-II), NetCTL-1.2 [46] and, mapped within the predicted epitopes.

Genome sequences of the SARS-CoV-2 from the World's different continents were obtained from NCBI Viruses [19] . The nucleotide sequence of different genes enables researchers to look into the virus behavior from a distance. Moreover, bioinformatics tools accelerate new alternatives of therapeutics before expensive, extensive research work. An antiviral peptide is one of such new alternatives [47] . Besides, antiviral peptides are successful in combating the SARS-CoV and MERS-CoV [48] [49] [50] . Therefore, with a similar kind of anticipation for the SARS-CoV-2, we have found three new antiviral peptides against the SARS-CoV-2 ( Figure   1a -c). Initially, the peptide sequences were interpreted as an antiviral peptide using AVPred [26] and Meta-iAVP [25] . The amino acid sequences of the predicted antiviral peptides were then mapped within the spike glycoprotein of the SARS-CoV-2. It is noteworthy that antiviral peptides Seq12, Seq12m, and Seq13m are analogous peptides of the spike glycoprotein AVPred is an antiviral peptide prediction server that is based on amino acid sequence features viz., motifs, alignment, amino acid composition, and physicochemical properties [26] .

Finally, the prediction of the antiviral peptide is made during 5-fold cross-validation using [25] . However, a most worrying concern of an antiviral peptide is its immunogenic profile [51] . A low immunogenic profile is the desired characteristic of an antiviral peptide because a low immunogenic profile reduces the chances of elimination by the host defense system [52] . Therefore, to intensify the immunogenic profile of the AVPs, we have performed epitope prediction using IEDB tools. Soon after epitope prediction, AVP sequences were mapped within the aligned sequence of the epitopes (only most frequently human alleles were chosen). Results showed that the epitopes do not have significant similarities with the AVPs ( Figure   S2 ). A few amino acid residues were apparently similar to the predicted epitopes. However, the binding postures (AVP-RBD) suggest these apparently similar amino acid residues were engaged with the AVP-RBD interactions. Further, the AVPs in this study can induce IL-4, an anti-inflammatory cytokine [53] . However, they are not capable of inducing IL-10 (a pro-inflammatory cytokine). Furthermore, AVPs were also non-allergenic as they do not have any known epitope for human IgE [32] . Other physicochemical properties viz., estimated half-life, instability index, water-solubility, theoretical IC 50 values, etc., are summarized in Table 1 . It is noteworthy that Seq12 and Seq12m have excellent PROB Score compared to the previously reported antiviral peptides, i.e., P7, P8, P9, and P10 ( Figure 2 ). These P7, P8, P9, and P10 are actually human ACE2 peptide mimics, and they block SARS-CoV-2 pulmonary cell infections [54] . In addition, we have found that the predicted IC 50 [27] new, meaning that it has acquired dozens of new mutations, and C-type is the current type, more contagious, and has more mutations [64] . The results depicted from this study suggest Seq12, Seq12m, and Seq13m might be effective to type A, B and C as well. As mentioned earlier, the interactions between RBD-ACE2 are essential for viral entry inside a human host cell [65] . A recent study has identified the molecular interactions between the RBD and ACE2 interactions [66] , which include six aromatic amino acid residues viz., TYR449, TYR453, PHE456, PHE486, TYR489, and TYR505, five polar uncharged amino acid residues viz., ASN487, GLN493, GLN498, THR500, and ASN501. And three non-polar aliphatic amino acid residues, viz., LEU455, GLY496, and GLY502. We utilized this information in molecular docking studies to zoom in on the molecular interactions of the AVPs-RBD complexes. However, there is an increasing amount of fear of mutations in the RBD, which would probably make all anti-COVID-19 efforts nil [67] . Indeed, it is true that if mutations are incorporated into the RBD, then therapeutics have to evolve accordingly.

Therefore, we have constructed a few in silico mutant models of the receptor-binding domain of the SARS-CoV-2, such as energy mutant (RBDe), evolutionary Figure S3 ). Therefore, we speculate that Seq12, Seq12m, and Seq13m might also be effective for the B1. Figure S4) . Furthermore, the RBDb/c/e are structurally distinguished from RBDm/RBD, and they can Results from molecular docking studies showed that AVP-RBD/RBDm complexes have thermodynamically favorable interactions (Table 4) . Besides, Seq12, Seq12m, and Seq13m

were engaged with nearly all critical amino acid residues of the RBD. Antiviral peptide of this study, i.e., Seq12, showed the best HADDOCK score of -111.2 kcal/mol followed by Seq13m (81.4 kcal/mol) and Seq12m (76.8 kcal/mol), suggesting AVPs were well-docked with the RBD (Figure 4) . Moreover, molecular docking results also suggest Seq12, Seq12m, units into a mature virus [70] . Therefore, protein-M also embraces importance as a potential therapeutic target. Results from our investigation indicate Seq12, Seq12m, and Seq13m are also participating in thermodynamically favorable interactions with the TM-1 region followed by TM-2 and TM-3, respectively (Figure 5a) . However, the molecular docking studies between the AVPs and active site of the viral RNA-dependent RNA polymerase (RdRp) were not thermodynamically favorable. Antiviral peptides usually rupture the viral capsid and eventually inhibit the viral replication cycles [69] . We speculate that the antiviral peptide, Seq12, Seq12m, and Seq13m could act as an anti-SARS-CoV-2 peptide by two possible mechanisms. Firstly, by inhibiting the RBD-ACE2 interaction. And secondly, by binding with M-protein followed by an eventual inhibition of the viral re-assembly/re-packaging. (Figure 8) . RMSF value is a measure of the average deviations of particular atoms or group of atoms from the initial reference structure [72] . Results showed that, in the case of the Seq13m-RBDm, RMSF fluctuation was high from about 138 th -158 th of the RBD.

Similarly, in the case of Seq13m-RBD, the RMSF value highly fluctuate from about 200 th to the rest of the C-alpha atoms ( Figure 8 ). However, only two crucial interacting amino acid residues, namely, PHE486 and TYR489, was remained within the fluctuating regions. And, the ΔG bind of Seq13m-RBDm and Seq13-RBD were also very good compared to the other RBD/RBDm-AVP complexes (Table 5 - Figure 9 . It showed that the R 2 of the regression model of the ΔG GBSA was 0.97 (p = 0.0003), suggesting ΔG bind is significantly correlated with the model. In addition, Figure 9a indicates that increasing the MD-Simulation time from 50 ns to 100 ns did not have an impact on the binding free energy.

However, in the case of ΔG PBSA the R 2 was 0.75 (p = 0.0256), although all variables were within 95% confidence intervals. This is because we have found that the ΔG PBSA calculated for 50 ns and 100 ns have high differences (Table S1 ). Furthermore, the MM-GBSA binding free energies per amino acid residues are summarized in Figure 10 . It showed that TYR489 is a crucial amino acid residue of the RBD because TYR489 is a common contributor to the best binding free energies of Seq12-RBD (-9.30 kcal/mol), Seq12m-RBD (-6.55 kcal/mol), properties of tyrosine may be the reason for this phenomenon. However, tyrosine did not participate in the top twenty amino acid residues in the case of Seq12m-RBDm. On the contrary, LEU493 was a poor contributor to the binding free energies of Seq12/Seq12m-RBDm interactions. In addition, ASN449 was another poor contributor in the case of Seq13m-RBDm interaction. Overall, the molecular details of AVP-RBD/RBDm interactions calculated using MM-GBSA provide evidence for the fact that AVPs were occupied the crucial amino acid residues of the RBD, which are necessary for the RBD-ACE2 interactions [65] .

Entropy of a thermodynamic system is the measure of the degree of freedom of the system where the translational (ΔS tr ), rotational (ΔS rot ), and vibrational (ΔS vib ) entropic terms determine the total entropy of the system [74, 75] . A change in total entropy correlates with the binding free energy, where the binding forces have to overcome the entropic effects during the binding process [75] . Therefore, a negative ΔS value indicates that the system becomes less disordered due to a decrease in the number of microstates for the proteinprotein complexes [43] . In particular, the entropic values of the AVP-RBD systems are negative, which explains the low distortion of the system (Table 7 ). In addition, the Seq13m-RBDm complex has the highest ΔS vib value (-0.43 kcal/mol), contributing to the increase of the phase space and binding energy of the system.

Antiviral peptides are a convenient alternative to conventional antiviral therapy. Now-a-days, AVPs are included in mainstream research against the SARS-CoV-2. In this study, peptide analogues of spike glycoprotein, i.e., Seq12, Seq12m, and Seq13m, showed antiviral properties against the SARS-CoV-2. These AVPs were derived from the RBD of the spike glycoprotein of the SARS-CoV-2. Seq12 and Seq12m showed negligible cytotoxicity, and they are also non-allergenic for humans. Besides, the predicted IC 50 values of the Seq12, Seq12m, and Seq13m are better than the anti-SARS-CoV-2 peptide P7, P9, P9, and P10.

Molecular dynamics simulation studies of the RBD/RBDm-Seq12 and RBD/RBDm-Seq12m

showed stable RMSD and RMSF throughout the complete MD-trajectory. Furthermore, the binding free energies, van der Waals interaction patterns calculated using MM-PB/GBSA are also in agreement with the molecular docking studies. The molecular docking and molecular dynamics simulation studies suggest Seq12, Seq12m, and Seq13m can block RBD, which is Furthermore, these AVPs can also interfere with viral membrane protein M. We speculate that these AVPs could eventually inhibit viral re-packaging cycles. Therefore, antiviral peptides, Seq12, Seq12m, and Seq13m could be helpful in the fight against the SARS-CoV-2.

Moreover, in the future, these AVPs could also help to develop anti-SARS-CoV-2 nasal spray. However, more studies are required before any clinical or diagnostic use. 

This research received no external funding.

Not applicable.

Not applicable.

The data presented in this study are available.

Council of Scientific and Industrial Research (CSIR), Govt. of India, New Delhi, India is also sincerely acknowledged by K.D. for Senior Research Fellowship (SRF), sanction letter no.

09/599(0082)/2019-EMR-I.

The authors declare no conflict of interest.

Digital technology and COVID-19

Diagnosis of SARS-CoV-2 infection and COVID-19: accuracy of signs and symptoms; molecular, antigen, and antibody tests; and routine laboratory markers

Airborne transmission of SARS-CoV-2: theoretical considerations and available evidence

How Indians responded to the Arogya Setu app?

Rapid, field-deployable nucleobase detection and identification using FnCas9, bioRxiv

Beclabuvir can inhibit the RNA-dependent RNA polymerase of newly emerged novel coronavirus (SARS-CoV-2)

Drug repurposing for coronavirus (COVID-19): in silico screening of known drugs against coronavirus 3CL hydrolase and protease enzymes

Tannins inhibit SARS-CoV-2 through binding with catalytic dyad residues of 3CLpro: An in silico approach with 19 structural different hydrolyzable tannins

Covid-19: Moderna vaccine is nearly 95% effective, trial involving high risk and elderly people shows

Evaluation of the immunogenicity of prime-boost vaccination with the replication-deficient viral vectored COVID-19 vaccine candidate ChAdOx1 nCoV-19

SARS-CoV-2 vaccines: status report

Antiviral peptides as promising therapeutic drugs

The Potential of Antiviral Peptides as COVID-19 Therapeutics

The SARS, MERS and novel coronavirus

epidemics, the newest and biggest global health threats: what lessons have we learned?

Goes de Jesus, Multiplex qPCR discriminates variants of concern to enhance global surveillance of SARS-CoV-2

COVID-19 Coronavirus spike protein analysis for synthetic vaccines, a peptidomimetic antagonist, and therapeutic drugs, and analysis of a proposed Achilles' heel conserved region to minimize probability of escape mutations and drug resistance

Structural basis of receptor recognition by SARS-CoV-2

NCBI viral genomes resource

I-TASSER server for protein 3D structure prediction

SWISS-MODEL and the Swiss-Pdb Viewer: an environment for comparative protein modeling

PROCHECK: a program to check the stereochemical quality of protein structures

MEME SUITE: tools for motif discovery and searching

Gapped BLAST and PSI-BLAST: a new generation of protein database search programs

Meta-iAVP: A sequence-based meta-predictor for improving the prediction of antiviral peptides using effective feature representation

AVPpred: collection and prediction of highly effective antiviral peptides

AVP-IC50Pred: Multiple machine learning techniques-based prediction of peptide antiviral activity in terms of half maximal inhibitory concentration (IC50)

PEP-FOLD3: faster de novo structure prediction for linear peptides in solution and in complex

GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers

Pep-Calc. com: a set of web utilities for the calculation of peptide and peptoid properties and automatic mass spectral peak assignment

ProtParam tool

AlgPred: prediction of allergenic proteins and mapping of IgE epitopes

In silico approach for predicting toxicity of peptides and proteins

HemoPred: a web server for predicting the hemolytic activity of peptides

A highly conserved cryptic epitope in the receptor binding domains of SARS-CoV-2 and SARS-CoV

FireProt: web server for automated design of thermostable proteins

The HADDOCK web server for data-driven biomolecular docking

Discovery studio visualizer

Paradoxical Bactericidal Effects of Hydrophobic Lung Surfactant Proteins and Their Peptide Mimics Using Liposome Molecular Trojan

Homology modeling and in silico design of novel and potential dual-acting inhibitors of human histone deacetylases HDAC5 and HDAC9 isozymes

CaFE: a tool for binding affinity prediction using end-point free energy methods

MMPBSA. py: an efficient program for end-state free energy calculations

HawkDock: a web server to predict and analyze the protein-protein complex based on computational docking and MM/GBSA

The immune epitope database (IEDB) 3.0

Large-scale validation of methods for cytotoxic T-lymphocyte epitope prediction

Peptide and peptide-based inhibitors of SARS-CoV-2 entry

MERS-CoV spike protein: a key target for antivirals

A novel peptide with potent and broad-spectrum antiviral activities against multiple respiratory viruses

Potential "biopeptidal" therapeutics for severe respiratory syndrome coronaviruses: a review of antiviral peptides, viral mechanisms, and prospective needs

Chimeric peptides: a new approach to enhancing the immunogenicity of peptides with low MHC class I affinity: application in antiviral vaccination

Antimicrobial peptides as therapeutic agents: opportunities and challenges

AIPpred: sequence-based prediction of anti-inflammatory peptides using random forest

Human ACE2 peptide-mimics block SARS-CoV-2 pulmonary cells infection

The integrin binding peptide, ATN-161, as a novel therapy for SARS-CoV-2 infection

TMPRSS2 and furin are both essential for proteolytic activation of SARS-CoV-2 in human airway cells

Investigation of ACE2 N-terminal fragments binding to SARS-CoV-2 Spike RBD, bioRxiv

A broad-spectrum virus-and host-targeting peptide against respiratory viruses including influenza virus and SARS-CoV-2

Furin inhibitors block SARS-CoV-2 spike protein cleavage to suppress virus production and cytopathic effects

A pan-coronavirus fusion inhibitor targeting the HR1 domain of human coronavirus spike

Inhibition of SARS-CoV-2 (previously 2019-nCoV) infection by a highly potent pancoronavirus fusion inhibitor targeting its spike protein that harbors a high capacity to mediate membrane fusion

Design of potent membrane fusion inhibitors against SARS-CoV-2, an emerging coronavirus with high fusogenic activity

In silico design of antiviral peptides targeting the spike protein of SARS-CoV-2

Phylogenetic network analysis of SARS-CoV-2 genomes

Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2

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

Implications of SARS-CoV-2 genetic diversity and mutations on pathogenicity of COVID-19 and biomedical interventions

Vaccine-induced immunity provides more robust heterotypic immunity than natural infection to emerging SARS-CoV-2 variants of concern

Spike protein recognition of mammalian ACE2 predicts the host range and an optimized ACE2 for SARS-CoV-2 infection

Sars-CoV-2 Envelope and Membrane proteins: differences from closely related proteins linked to cross-species transmission

A molecular modeling approach to identify effective antiviral phytochemicals against the main protease of SARS-CoV-2

Automatic identification of mobile and rigid substructures in molecular dynamics simulations and fractional structural fluctuation analysis

Assessing the performance of the MM/PBSA and MM/GBSA methods. 6. Capability to predict protein-protein binding free energies and re-rank binding poses generated by protein-protein docking

The meanings of entropy

Insights into protein-ligand interactions: mechanisms, models, and methods

Hydrogen bonding in biological macromolecules

Hydrogen bonding in biological structures

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.