In Silico Analysis of Plant Flavonoids as Potential Inhibitors of Newcastle Disease Virus V Protein

Round 1

Reviewer 1 Report

General Comments
1. The experiment was good and followed the scientific method.
2. The manuscript is good and scientifically presented
3. The language used is good and followed a scientific writing style. To the best of my knowledge, it is free of grammatical errors.
4. Therefore, I strongly recommend it for publication

Author Response

Reviewer 1

General Comments
1. The experiment was good and followed the scientific method.
2. The manuscript is good and scientifically presented
3. The language used is good and followed a scientific writing style. To the best of my knowledge, it is free of grammatical errors.
4. Therefore, I strongly recommend it for publication

Answer: The comments from reviewer 1 are really appreciated and are self-explanatory. Editor should keep these comments under consideration while reviewing manuscript.

Reviewer 2 Report

In their article In-silico analysis of Plant flavonoids as potential inhibitors of Newcastle Disease Virus V Protein, Sarwar et. al claim that their results prove the potential of flavonoids as V protein inhibitors. Not only is this a strong statement, but it is also incorrect. 

Without experimental validation, the docking results obtained by the group are not useful. This is based on multiple different reasons:

1. None of the screened compounds have been tested for antiviral activity or even in-vitro binding affinities
2. The lack of positive or negative controls severely impacts the reliability of the results
3. Especially for flavonoids, the possibility of pan-assay interference compounds (PAINS) needs to be considered. This has been found to be a problem before: Jiménez-Avalos, G., Vargas-Ruiz, A.P., Delgado-Pease, N.E. et al. Comprehensive virtual screening of 4.8 k flavonoids reveals novel insights into allosteric inhibition of SARS-CoV-2 M^PRO. Sci Rep 11, 15452 (2021). https://doi.org/10.1038/s41598-021-94951-6. Therefore, results obtained solely from docking cannot be considered significant and need to be backed up by experimental data. 

For these reasons, I recommend rejecting the presented article on the basis that the results might be overinterpreted without experimental backup.

Author Response

Reviewer 2

In their article In-silico analysis of Plant flavonoids as potential inhibitors of Newcastle Disease Virus V Protein, Sarwar et al., claim that their results prove the potential of flavonoids as V protein inhibitors. Not only is this a strong statement, but also it is also incorrect.

Answer: We have toned down the statement throughout the manuscript.

Without experimental validation, the docking results obtained by the group are not useful. This is based on multiple different reasons:

1. None of the screened compounds have been tested for antiviral activity or even in-vitro binding affinities. The lack of positive or negative controls severely impacts the reliability of the results. Especially for flavonoids, the possibility of pan-assay interference compounds (PAINS) needs to be considered. This has been found to be a problem before: Jiménez-Avalos, G., Vargas-Ruiz, A.P., Delgado-Pease, N.E. et al.Comprehensive virtual screening of 4.8 k flavonoids reveals novel insights into allosteric inhibition of SARS-CoV-2 M^PRO. Sci Rep 11, 15452 (2021). https://doi.org/10.1038/s41598-021-94951-6. Therefore, results obtained solely from docking cannot be considered significant and need to be backed up by experimental data. 

For these reasons, I recommend rejecting the presented article on the basis that the results might be over interpreted without experimental backup.

Answer: This manuscript does not claim any in-vivo or in-vitro assays. As title of the manuscript suggests, these are only In-silico experiments. However, seeking the in-vivo validation is also necessary. We believe that reviewer is right in asking the experimental verification of flavonoids interaction with viral proteins In the future several In-vitro or In-vivo experiments will also be performed. Currently, there are number of manuscripts which are recently published based on In-silico results. Such studies provide the basis of future experiments on drug discoveries.

Reviewer 3 Report

This research conducted a virtual screening towards Newcastle disease virus V protein. The protein modeling and docking were performed and 10 flavonoids were selected as potential inhibitors. The method is fair and the compounds found in this simulation are interesting. However, several points should be clarified for publication.

Major point
1. Is there any known small molecule inhibitors or natural ligand for NDV V protein or homologous proteins? If exist, please refer to some in the introduction.
2. For protein modeling, please clarify the protein similarity towards the template protein (identity or similarity).
3. Please clarify the reason or criteria to select this number of compounds (10) as potential inhibitors.
4. Similar to point 1, if there are any known small molecule inhibitors or natural ligands for NDV V protein, please compare the docking pose and score to your in silico hits.
5. Please confirm that the interacting residues are conserved among strains of NDV. If not, the virus can easily avoid these interactions, and that should be mentioned.


Minor point
1. Use consistent capitalization for the paper title.
2. Lines 46-47. It hinders interferon signaling by steering signal transducer and transcription activator of (STAT1) ….：Remove “of”.
3. Line 163. Should “Nile et al, 2020” be a reference pattern like [47]?
4. Lines 166-167. In vitro and in vivo should be italic.
5. “antiviral” and “anti-viral” are not unified.

Author Response

Reviewer 3

This research conducted a virtual screening towards Newcastle disease virus V protein. The protein modeling and docking were performed and 10 flavonoids were selected as potential inhibitors. The method is fair and the compounds found in this simulation are interesting. However, several points should be clarified for publication.

Major points;

1. Is there any known small molecule inhibitors or natural ligand for NDV V protein or homologous proteins? If exist, please refer to some in the introduction.

Answer: Currently, as per literature search we don’t see any example of small molecule inhibitors for NDV-V protein. Therefore at this moment we cannot present conclusive evidence in the introduction part.

1. For protein modeling, please clarify the protein similarity towards the template protein (identity or similarity).
   
   

Answer: We have used Swiss model for protein modeling and evaluation. Sequence similarity is one factor. Sometimes we use less sequence similarity for the sake of assay. The criterion of modeling from Swiss model was based on Sequence homology, Sequence coverage, Energy plots and Z score. We have to choose the best model by considering values of all four factors. Although at sequence level maximum sequence percentage is required. But for a good template this not enough because a protein bigger in sequence lose it's percentage identity yet there many residues available for act as template in model building.

 

3. Please clarify the reason or criteria to select this number of compounds (10) as potential inhibitors.

Answer: These 10 compounds were selected on the basis of best docking score. The docking score table is presented in the manuscript as Table 1.

4. Similar to point 1, if there are any known small molecule inhibitors or natural ligands for NDV V protein, please compare the docking pose and score to your in silico hits.

Answer: This question has already been addressed.

5. Please confirm that the interacting residues are conserved among strains of NDV. If not, the virus can easily avoid these interactions, and that should be mentioned.

   
   

Answer: These ten flavonoids interact with different residues of V protein of NDV. But there are certain interactions, which are conserved for certain flavonoids e.g., Gln 63 and Gln 160 of V protein interacts with Flavonoid 4, 6 and 8 but we selected best flavonoids on the basis of best docking score. It is possible that virus can avoid these interactions but there are multiple interactions between flavonoid and V protein and potentially these flavonoids could be useful in devising resistant strategies against NDV.

Minor point
1. Use consistent capitalization for the paper title.
2. Lines 46-47. It hinders interferon signaling by steering signal transducer and transcription activator of (STAT1) ….：Remove “of”.
3. Line 163. Should “Nile et al, 2020” be a reference pattern like [47]?
4. Lines 166-167. In vitro and in vivo should be italic.
5. “antiviral” and “anti-viral” are not unified.

Answer: All these 5 points are addressed in the manuscript.

Reviewer 4 Report

The topic presented in the manuscript is interesting. But, I recommend rejecting the manuscript due to the following two major reasons:

1. Research was not conducted correctly, e.g., (i) there is no justification of the template selection for the homology modeling; (ii) there is no explanation about how the best model can be assigned as "the best model"; and (iii) there is no validation for the molecular docking used in the research.  
2. Results of the molecular docking simulations per se are no longer acceptable to come to such a conclusion "We conclude that these flavonoids can bind to the V protein 173 of NDV and can inhibit the activity of this protein; so all these ten flavonoids possess 174 potential anti-NDV activity." Further molecular dynamics simulations with binding energy calculations are highly suggested to come to such a conclusion. 

Author Response

Reviewer 4

The topic presented in the manuscript is interesting. But, I recommend rejecting the manuscript due to the following two major reasons:

1. Research was not conducted correctly, e.g., (i) there is no justification of the template selection for the homology modeling; (ii) there is no explanation about how the best model can be assigned as "the best model"; and (iii) there is no validation for the molecular docking used in the research.  

 

Answer: When experimentally determined protein 3D structure is not available, homology modeling plays a cost-effective role in structure-based applications and the characterization of protein properties and functions. Template selection is fully justified. The Proper template was selected for the target sequence through BLAST search using target protein sequence as the query and PDB as database. It was tried hard that the selected template is structurally and biologically most relevant. For this, multiple template information, refined alignments in nontrivial cases, selection of segments for loop modeling, including cofactors and ligands in the model were optimally combined. As per Critical Assessment of techniques for protein Structure Prediction (CASP), the accurateness of the model was assessed in terms of both global and local structure accuracy. Global structure accuracy estimation was evaluated by the global structure score value while evaluating residue-wise accuracy ensured local structure accuracy. Template based selection offers high-quality models that are comparable with low-resolution X-ray crystallography or medium-resolution NMR solution structures.

 

2. Results of the molecular docking simulations per se are no longer acceptable to come to such a conclusion "We conclude that these flavonoids can bind to the V protein 173 of NDV and can inhibit the activity of this protein; so all these ten flavonoids possess 174 potential anti-NDV activity." Further molecular dynamics simulations with binding energy calculations are highly suggested to come to such a conclusion. 

 

Answer: The results of docking simulations are widely acceptable. However, we agreed with the reviewer that conclusive statements could only be provided if interactions of ligand and proteins are confirmed. Therefore, we have toned down the statements in the results section.

                                                                                                

Reviewer 5 Report

Line 46-47: It hinders interferon signaling by steering signal transducer and transcription activator of (STAT1) for its deprivation by the carboxyl end domain.

Consider expatiating on the above to include the full meaning of STAT1 and also note that with STAT1 in the bracket, the sentence does not flow.

The abbreviation MOE is not introduced for the first time. some readers might struggle to figure out what is MOE

 

Author Response

Reviewer 5

1. Line 46-47: It hinders interferon signaling by steering signal transducer and transcription activator of (STAT1) for its deprivation by the carboxyl end domain.
2. Consider expatiating on the above to include the full meaning of STAT1 and also note that with STAT1 in the bracket, the sentence does not flow.
3. The abbreviation MOE is not introduced for the first time. Some readers might struggle to figure out what is MOE

Answer: All these 3 points have been addressed/corrected in the manuscript.

Round 2

Reviewer 2 Report

As I have said before, I fear that information solely based on in-silico data does not provide sufficient scientific information for these types of compounds. Without any form of experimental validation, obtained hits are a mere speculation.

Author Response

Molecular modeling techniques have been playing an increasingly crucial role in the search for new drugs and their optimization, in every area of drug design. This is clear from the increasing number of literature sources concerned with in-silico approaches and the boost in the number of molecular modeling packages. There are number of reports e.g., Rameshkumar et al., 2021. (https://doi.org/10.1016/j.sjbs.2020.10.028). Simsek et al., 2021. (https://doi.org/10.1016/j.jmgm.2021.108038). Teli et al., 2021. (https://doi.org/10.3389/fmolb.2020.599079) published in peer-reviewed journals based on in-silico results. There are recent reports about SARS-CoV-2. Researchers have also published similar studies about HIV and many other viruses. The application of in-silico technologies to antiviral research has already led to the design of several compounds approved by the US Food and Drug Administration and now used in clinical therapy (Artese et al., 2020 doi: 10.1016/j.drup.2020.100721) and (Congreve et al., 2005, DOI: 10.1016/S1359-6446(05)03484-7). We tried to run simulations with the model and ligand structures presented in this manuscript but unfortunately Schrodinger software that, we used didn't support it and we were unable to run simulations. If respected reviewer insists that we should run simulations then we have to construct protein models on a new modeler and all the analysis presented in this manuscript have to be changed. 

Reviewer 4 Report

1. The authors have responded some of the 1st comments properly, but there is one critical comment remain without response, i.e., "There is no validation for the molecular docking used in the research".
2. The authors have toned done the statements but in this context I do not agree with the response "The results of docking simulations are widely acceptable." Indeed, docking simulations are widely acceptable, but it needs proper validations depends on the purpose of the simulations (see Beware of docking! - PubMed (nih.gov)). And, for the purpose to predict the biological activity (in this manuscript "potential 20 inhibitors of the V protein of Newcastle disease virus"), retrospective structure-based virtual screening campaigns (SBVS) are required. These retrospective SBVS campaigns have been introduced in 2006 (see Benchmarking sets for molecular docking - PubMed (nih.gov)) and widely acceptable since. Moreover, there is no response on the suggestion "Further molecular dynamics simulations with binding energy calculations are highly suggested to come to such a conclusion." 

Therefore the manuscript is still not acceptable to be published in the current form. More simulations or data are required before it can be considered to be published, i.e.:

• Validation of the molecular docking simulations used in the manuscript, and
• Short molecular dynamics simulations to examine the protein-ligand complex stability (see Exploring the stability of ligand binding modes to proteins by molecular dynamics simulations - PubMed (nih.gov)) 

Author Response

1. The authors have responded some of the 1st comments properly, but there is one critical comment remain without response, i.e., "There is no validation for the molecular docking used in the research".
2. The authors have toned done the statements but in this context I do not agree with the response "The results of docking simulations are widely acceptable." Indeed, docking simulations are widely acceptable, but it needs proper validations depends on the purpose of the simulations (see Beware of docking! - PubMed (nih.gov)). And, for the purpose to predict the biological activity (in this manuscript "potential 20 inhibitors of the V protein of Newcastle disease virus"), retrospective structure-based virtual screening campaigns (SBVS) are required. These retrospective SBVS campaigns have been introduced in 2006 (see Benchmarking sets for molecular docking - PubMed (nih.gov)) and widely acceptable since. Moreover, there is no response on the suggestion "Further molecular dynamics simulations with binding energy calculations are highly suggested to come to such a conclusion." 

Therefore the manuscript is still not acceptable to be published in the current form. More simulations or data are required before it can be considered to be published, i.e.:

• Validation of the molecular docking simulations used in the manuscript, and
• Short molecular dynamics simulations to examine the protein-ligand complex stability (see Exploring the stability of ligand binding modes to proteins by molecular dynamics simulations - PubMed (nih.gov)) 

 

Answer:

• This manuscript does not claim any in-vivo or in-vitro As title of the manuscript suggests, these are only In-silico experiments. However, seeking the in-vivo validation is also necessary. We believe that reviewer is right in asking the experimental verification of flavonoids interaction with viral proteins In the future several In-vitro or In-vivo experiments will also be performed. Currently, there are number of manuscripts which are recently published based on In-silico results. Molecular modeling techniques have been playing an increasingly crucial role in the search for new drugs and their optimization, in every area of drug design. This is clear from the increasing number of literature sources concerned with in-silico approaches and the boost in the number of molecular modeling packages. There are number of reports e.g., Rameshkumar et al., 2021. (https://doi.org/10.1016/j.sjbs.2020.10.028). Simsek et al., 2021. (https://doi.org/10.1016/j.jmgm.2021.108038). Teli et al., 2021. (https://doi.org/10.3389/fmolb.2020.599079) published in peer-reviewed journals based on in-silico results. There are recent reports about SARS-CoV-2. Researchers have also published similar studies about HIV and many other viruses. The application of in-silico technologies to antiviral research has already led to the design of several compounds approved by the US Food and Drug Administration and now used in clinical therapy (Artese et al., 2020 doi: 10.1016/j.drup.2020.100721) and (Congreve et al., 2005, DOI: 10.1016/S1359-6446(05)03484-7).
• We tried to run simulations with the model and ligand structures presented in this manuscript but unfortunately Schrodinger software that, we used didn't support it and we were unable to run simulations. If respected reviewer insists that we should run simulations then we have to construct protein models on a new modeler and all the analysis presented in this manuscript have to be changed.