Review Reports

Round 1

Reviewer 1 Report

The authors of the paper 'Pharmacophore-based virtual screening, quantum mechanics  calculations, and molecular dynamics simulation approaches  identified potential natural antiviral drug candidates against MERS-CoV S1-NTD' presented a well-designed project of the virtual search for novel MERS-CoV S1-NTD inhibitor. The research has good structure and design and could be published in Molecules provided that language errors are removed and one section corrected that seems wrong. 

My only methodological reservation is QM-calculations of the geometry of the ligands (and perhaps associated with it FMO analysis). One can see from the figures depicting 3D structures of the ligands that they are not optimized (distorted rings, elongated bonds). I have no idea how any chemist or biologist could think that these structures are optimized? Especially while MM minimization should have already given a pretty good approximation of the geometries. These geometries (different conformers) should be optimized and frequency analysis should be used to validate if minima were obtained (for the multiple ring ligands there will be several conformers with close energy and these conformational changes will influence DG and structure of HOMO and  LUMO orbitals).

I also suggest comparing MD trajectories of the studied ligand with the crystal structure and the ligand co-crystalized. The structure of the protein-inhibitor complex will always change in solution and during MD so comparing RMSD to static crystal structure is ok but may actually lead to belive that significant changes occurred (RMSD of 3A is quite a big difference). 

 

Only this section seems to be wrong - the rest can be corrected, figures should be of better quality and so on. However, I do not have any bigger reservations about the rest of the study. My detailed remarks are provided in the attached file. 

I also strongly suggest that representative complexes of protein-inhibitors (for example dominant cluster from cluster analysis of trajectories) are provided in SI or other data repository so it can be examined by other authors and used to develop new drugs. 

Author Response

“Response to Reviewers Comments”

Referee: 1

Comments to the Author:

The authors of the paper 'Pharmacophore-based virtual screening, quantum mechanics calculations and molecular dynamics simulation approaches identified potential natural antiviral drug candidates against MERS-CoV S1-NTD' presented a well-designed project of the virtual search for novel MERS-CoV S1-NTD inhibitor. The research has good structure and design and could be published in Molecules provided that language errors are removed, and one section corrected that seems wrong. 

Comment: My only methodological reservation is QM-calculations of the geometry of the ligands (and perhaps associated with it FMO analysis). One can see from the figures depicting 3D structures of the ligands that they are not optimized (distorted rings, elongated bonds). I have no idea how any chemist or biologist could think that these structures are optimized? Especially while MM minimization should have already given a pretty good approximation of the geometries. These geometries (different conformers) should be optimized, and frequency analysis should be used to validate if minima were obtained (for the multiple ring ligands there will be several conformers with close energy and these conformational changes will influence DG and structure of HOMO and LUMO orbitals).

Reply: Thank you for your comment. We have improved the section by providing sufficient rationale regarding the geometry optimization process. Please find it in the Supplementary file as text format (renamed as Geometry). During the analysis process, we have also thought like you, but after different confirmation analyses, we found that the geometry has optimized, and the optimized structure has been utilized for FMO calculation. The bond angles, bond lengths (Bohr, angstroms) and torsional angles optimized during the process have also provided in the Supplementary file section.

Comment: I also suggest comparing MD trajectories of the studied ligand with the crystal structure and the ligand co-crystalized. The structure of the protein-inhibitor complex will always change in solution and during MD so comparing RMSD to static crystal structure is ok but may actually lead to belive that significant changes occurred (RMSD of 3A is quite a big difference). 

Reply: Thank you for your suggestion. Previously, we performed a MD simulation of ligand (folic acid) in complex with the crystal structure of the protein. The selected ligand and the co-crystal structure RMSD have combinedly analyzed and compared that has provided in the supplementary figure. I am also agreeing with your concern that the RMSD of 3A is quite a big difference, but in the case of the study the fluctuation has found in average 2.5 A.

Comment: Only this section seems to be wrong - the rest can be corrected; figures should be of better quality and so on. However, I do not have any bigger reservations about the rest of the study. My detailed remarks are provided in the attached file. 

Reply: Thank you for your comment. We have corrected the manuscript according to the attached file you provided.

Comment: I also strongly suggest that representative complexes of protein-inhibitors (for example dominant cluster from cluster analysis of trajectories) are provided in SI or another data repository so it can be examined by other authors and used to develop new drugs. 

Reply: Thank you for your suggestion. We have added the complexes of protein-inhibitors (for example dominant cluster from cluster analysis of trajectories) in the supplementary file.

Reviewer 2 Report

In this computational work the Authors have tried to find the compounds with the activity against MERS-CoV S1-NTD. Though the study is rather routine, it is also acceptable as the methods have been used properly (in most cases). There are, however, some major and minor issues listed below that must be fixed before proceeding further.

The abstract is significantly too long. It should summarize the study and not be the repeat of introduction.

Line 84, the information whether there are any compounds with clinically (or at least in vitro) proven inhibitory activity against S1-NTD must be provided.

Lines 113-122, do the compounds with the experimentally proved activity (see my previous comment) fit into this model?

Line 126, this sentence should be rewritten as it is grammatically incorrect.

The abbreviations such as HBD, HBA, NI, AR should be defined in the main text, not only in the figure caption.

Lines 135-140, this should be moved to the Methods section.

Line 146, how did you know that those compounds act through S1-NTD and not S1-CTD? If there is no such information those compound should not be used as decoys.

A minor editorial issue: the text should be justified.

Lines 179-182, a reference is needed here to support this statement.

Line 182, why factor with a capital “F”?

Figure 3 is not very informative, it should be moved to ESI.

Table 1, please add the structures of the compounds.

Line 274, this is not true. Geometry optimization can be done also on other levels of theory, i.e. molecular mechanics, not only QM. Besides, not only computational biologists but also chemists and other academics and researchers.

Line 278, this is not true. Actually, one should take into consideration not energy but thermodynamic properties (i.e. free enthalpy).

Line 280, this is a VERY small basis set, it should be 6-311++G(d,p) in this case.

The conformational search should be done prior to geometry optimization.

Line 746, it should be “ps”

 

Author Response

Reviewer: 2

 

Comments to the Author:

In this computational work, the Authors have tried to find the compounds with the activity against MERS-CoV S1-NTD. Though the study is rather routine, it is also acceptable as the methods have been used properly (in most cases). There are, however, some major and minor issues listed below that must be fixed before proceeding further.

 

Comment: The abstract is significantly too long. It should summarize the study and not be the repeat of the introduction.

Reply: Thank you for your suggestion. The section has been improved and summarized according to your instruction.

 

Comment: Line 84, the information whether there are any compounds with clinically (or at least in vitro) proven inhibitory activity against S1-NTD must be provided.

Reply: Thank you for your Co. Previously, folic acid has shown activity against NTD from the mammalian expression medium, but closely related chemical analog-like to folic acid and their activity against MERS-CoV infection cycle remains unknown. Please find it in the introduction section.

 

Comment: Lines 113-122 do the compounds with the experimentally proved activity (see my previous comment) fit into this model?

Reply: Thank you for your comment. The experimentally validated 3D X-ray crystal structure of the protein in complex with the folic acid has been used to generate the model. Therefore, it was not necessary to see the fit value of the compound, as the compound has used o generated the main pharmacophore model in this study.

 

Comment: Line 126, this sentence should be rewritten as it is grammatically incorrect.

Reply: Thank you for your comment. The sentence has been rewritten and improved.

 

Comment: The abbreviations such as HBD, HBA, NI, AR should be defined in the main text, not only in the figure caption.

Reply: Thank you for your suggestion. The abbreviation has been provided in the main text.

 

Comment: Lines 135-140, this should be moved to the Methods section.

Reply: Thank you for your suggestion. The line has been moved in the method section.

 

Comment: Line 146, how did you know that those compounds act through S1-NTD and not S1-CTD? If there is no such information those compounds should not be used as decoys.

Reply: Thank you for your comment. Initially, twelve experimentally active compounds against MERS-CoV S1-NTD protein have been identified and retrieved from the ChEMBL database. The compounds then submitted in the DUDE decoy database and they provided the correspondence decoy compound against the active compound. As we submitted the active compound of S1-NTD, the DUDE decoy database provides decoy compounds only for S1-NTD. Please find more details in method section.

 

A minor editorial issue:

 

Comment: the text should be justified.

Reply: The text has been justified.

Comment: Lines 179-182, a reference is needed here to support this statement.

Reply: The reference has been included.

 

Comment: Line 182, why factor with a capital “F”?

Reply: Corrected.

 

Comment: Table 1, please add the structures of the compounds.

Reply: The structure of the compounds has been added in the Table 1.

 

Comment: Line 274, this is not true. Geometry optimization can be done also on other levels of theory, i.e., molecular mechanics, not only QM. Besides, not only computational biologists but also chemists and other academics and researchers.

Reply: Thank you for your suggestion. The sentence has been improved according to your suggestion.

 

Comment: Line 278, this is not true. One should take into consideration not energy but thermodynamic properties (i.e., free enthalpy).

Reply: Thank you for your suggestion. The section has corrected.

 

Comment: Line 280, this is a VERY small basis set, it should be 6-311++G(d,p) in this case.

Reply: Thank you for your suggestion. Corrected

 

Comment: The conformational search should be done prior to geometry optimization.

Reply: Thank you for your suggestion. We have search confirmation for each selected ligand and after that, the geometry has optimized. Please find it in the supplementary file.

 

Comment: Line 746, it should be “ps”

Reply: Thank you for your suggestion. Corrected.

Round 2

Reviewer 2 Report

The Authors have corrected their manuscript and it can now be published as it is.