Response to the Editor and Reviewers

Manuscript ID: PONE-D-26-27333

Title: Computational Identification and Evaluation of Adenium obesum Phytochemicals as Potential EGFR Inhibitors for Targeted Therapy Against Non-Small-Cell Lung Cancer

Dear Editor and Reviewers,

We sincerely thank the Academic Editor and all reviewers for their valuable time and insightful comments. We have revised our manuscript thoroughly and addressed all comments point by point below. All changes have been incorporated into the revised version, with tracked changes indicated. We believe these revisions have significantly improved the quality and clarity of our manuscript. All the revisions are highlighted by yellow color.

Response to the Editor

We sincerely thank the Editor for the careful evaluation of our manuscript and for providing constructive guidance regarding the major issues that required attention. We fully acknowledge the concerns raised and have substantially revised the manuscript to improve the scientific rigor, balance, methodological transparency, and clinical relevance of the study.

In response to the Editor's comments, we have made the following major revisions:

1. Experimental validation and interpretation of computational findings: We have explicitly acknowledged throughout the revised manuscript that the present study is entirely computational and that the findings are hypothesis-generating. Claims suggesting confirmed EGFR inhibition or therapeutic efficacy have been moderated accordingly. Cardenolide is now described as a computationally prioritized candidate or potential EGFR-targeting compound, and the need for subsequent biochemical, cellular, and in vivo experimental validation has been clearly stated.

2. Docking-score interpretation: We have revised the molecular docking Results and Discussion to clarify that docking scores represent predicted binding characteristics and cannot independently establish binding affinity, inhibitory potency, or therapeutic efficacy. Comparisons with approved EGFR inhibitors have been reframed as computational observations rather than evidence of superior inhibitory activity.

3. Clinically relevant EGFR mutations: To improve the translational relevance of the study, we performed additional comparative molecular docking of all 16 investigated Adenium obesum phytochemicals against EGFR structures representing wild-type EGFR (PDB ID: 1M17), L858R/T790M EGFR (PDB ID: 4I22), and L858R/T790M/C797S EGFR (PDB ID: 6LUD). The revised manuscript now presents and discusses the compound-dependent differences in predicted docking behavior across these structural backgrounds. We have also clearly acknowledged that these analyses do not establish mutant EGFR inhibition or the ability to overcome clinically established EGFR resistance mechanisms.

4. Rationale for Cardenolide prioritization: We have clarified that Cardenolide was not selected solely on the basis of its docking score. Although SCHEMBL5933906 and Echujin demonstrated comparable or more favorable predicted docking scores, Cardenolide was computationally prioritized based on the overall balance of its predicted docking characteristics, physicochemical properties, drug-likeness, pharmacokinetic profile, and toxicity-related predictions. We have also explicitly acknowledged that SCHEMBL5933906 and Echujin were not subjected to molecular dynamics simulations and therefore that MD results were not used as a comparative criterion for selecting Cardenolide over these compounds.

5. Pharmacokinetic limitations and Lipinski violations: The revised manuscript now provides a more balanced discussion of the predicted drug-likeness and pharmacokinetic limitations of the investigated phytochemicals. In particular, we discuss the multiple predicted Lipinski rule violations, high molecular weight, elevated TPSA, and low predicted gastrointestinal absorption observed for several highly ranked glycosylated phytochemicals. We have also clarified that Cardenolide itself should not be considered uniformly optimal, particularly in view of its predicted lipophilicity, and that all ADME and toxicity predictions require experimental validation.

6. Limitations of the computational approach: The Discussion and Conclusion have been revised to provide a more balanced interpretation of the findings. The limitations of molecular docking, molecular dynamics, ADME prediction, toxicity prediction, and other computational approaches have been more explicitly acknowledged, and the need for experimental validation has been emphasized.

7. Language and manuscript organization: The manuscript has undergone extensive language editing to correct grammatical and typographical errors. The Introduction has been condensed to remove repetitive and unnecessary details while retaining the essential background, research gap, rationale, and objectives.

8. Methodological transparency: The Methods section has been revised to provide more consistent information regarding software versions, access dates, docking parameters, receptor structures, and grid-box settings. The grid-box parameters for the wild-type and mutation-containing EGFR structures are now reported in a consolidated table to facilitate reproducibility.

9. Molecular interaction visualization: The molecular docking figures have been improved by providing higher-resolution interaction images and informative two-dimensional protein–ligand interaction diagrams to facilitate clearer interpretation of the predicted binding interactions.

10. Comparison with previously reported natural EGFR-related compounds: The Discussion has been expanded to briefly contextualize the present findings with previously investigated natural products, including curcumin, epigallocatechin-3-gallate (EGCG), quercetin, resveratrol, berberine, and genistein. This comparison helps establish the novelty of investigating Adenium obesum phytochemicals through an integrated computational framework while maintaining appropriate caution regarding the distinction between EGFR-related activity and experimentally confirmed direct EGFR inhibition.

We believe that these comprehensive revisions have substantially strengthened the manuscript and have addressed the Editor's and Reviewer's concerns regarding the interpretation, methodological rigor, translational relevance, and presentation of the study. We have also carefully reviewed the revised manuscript to ensure that the conclusions are appropriately aligned with the computational nature of the evidence and do not exceed the limitations of the study.

We sincerely appreciate the Editor's and Reviewer's constructive comments, which have helped us improve the quality, clarity, and scientific rigor of the manuscript.

Response to the Reviewers

Reviewer #1

Major Comments

Comment 1: The study is entirely computational, and the conclusions regarding Cardenolide as a promising EGFR inhibitor are not supported by in vitro or in vivo validation.

Author Response:

We sincerely thank the reviewer for this important and constructive comment. We fully agree that the absence of experimental validation is an important limitation of the present study and that the computational findings should not be interpreted as evidence of experimentally confirmed EGFR inhibitory activity or therapeutic efficacy. The objective of our study was to computationally prioritize phytochemicals from Adenium obesum based on their predicted EGFR-binding characteristics, molecular interactions, structural stability, physicochemical properties, ADMET profiles, and toxicity predictions. We acknowledge that these computational approaches are hypothesis-generating and cannot substitute for direct biochemical, cellular, or in vivo validation.

In response to the reviewer's concern, we have carefully revised the manuscript to ensure that the conclusions are appropriately aligned with the evidence generated by the computational analyses. In particular, we have removed or modified statements that could be interpreted as claiming that Cardenolide is a confirmed EGFR inhibitor or that it possesses demonstrated therapeutic efficacy against NSCLC. Throughout the revised manuscript, we now consistently refer to Cardenolide as a "computationally prioritized candidate," "putative EGFR-binding candidate," or "potential EGFR-targeting compound," where appropriate.

We have also explicitly acknowledged that the present study is entirely computational and that the findings should be considered hypothesis-generating. We now clearly state that experimental validation is required to determine whether the predicted interactions translate into actual EGFR inhibitory activity and anticancer efficacy. Specifically, we have highlighted the need for biochemical assays, including direct EGFR kinase inhibition studies, followed by cellular validation in relevant NSCLC models. Further experimental studies will also be necessary to establish the pharmacokinetic properties, mechanism of action, and safety profile of Cardenolide.

The following major revisions have been made:

1. Abstract: The statement identifying Cardenolide as a "potential EGFR inhibitor" has been revised to describe it as a "computationally prioritized candidate with favorable predicted EGFR-binding characteristics, structural stability, and physicochemical and toxicity profiles." We have also explicitly stated that the study is entirely computational, that the findings are hypothesis-generating, and that biochemical and cellular experimental validation is required.

2. Introduction: We have revised the study rationale and objectives to clarify that the computational workflow is designed to prioritize candidate EGFR-targeting compounds rather than establish experimentally confirmed inhibitory activity. We now explicitly state that the computational findings provide a basis for selecting candidates for subsequent biochemical and cellular validation.

3. Conclusion: The conclusion has been substantially revised to clarify that the predicted docking scores, molecular interactions, molecular dynamics stability, and in silico ADMET and toxicity profiles do not establish experimentally confirmed EGFR inhibitory activity, anticancer efficacy, or therapeutic safety. Cardenolide is now presented as a computationally prioritized candidate for further experimental investigation, rather than as a confirmed EGFR inhibitor.

4. Throughout the manuscript: We have systematically reviewed the terminology and revised overstatements where necessary. Claims implying confirmed EGFR inhibition, therapeutic efficacy, or safety have been replaced with appropriately cautious terminology such as "predicted EGFR-binding potential," "computationally prioritized candidate," and "candidate EGFR-targeting compound." Similarly, statements concerning ATP-site binding are now framed as predicted structural compatibility with the EGFR ATP-binding pocket rather than experimentally established ATP-competitive inhibition.

We believe these revisions have substantially improved the scientific rigor and transparency of the manuscript by ensuring that our conclusions accurately reflect the scope and limitations of the computational evidence. We are grateful to the reviewer for highlighting this important issue, which has helped us to better distinguish between computational prioritization and experimentally validated biological activity and to present Cardenolide as a testable candidate for future biochemical and cellular investigation rather than as a confirmed EGFR inhibitor.

Comment 2: The manuscript overstates the superiority of several phytochemicals over approved EGFR inhibitors based solely on docking scores, whereas docking energies alone are insufficient to establish inhibitory potency or therapeutic efficacy.

Author Response:

We sincerely thank the reviewer for this important observation. We fully agree that docking scores are model-dependent computational estimates and should not be interpreted as experimentally measured binding affinities, binding free energies, inhibitory potencies, or therapeutic efficacy. We acknowledge that our original wording could have inadvertently overstated the significance of the more favorable docking scores obtained by several Adenium obesum phytochemicals relative to the reference EGFR-targeting compounds.

In response to this comment, we have substantially revised the presentation and interpretation of the molecular docking results throughout the manuscript. We no longer describe the phytochemicals as having experimentally "higher affinity" or as "outperforming" approved EGFR-targeting drugs based solely on docking scores. Instead, the docking results are now presented as relative computational ranking metrics within the applied docking protocol, with more negative scores indicating more favorable predicted docking scores under the AutoDock Vina scoring function.

Specifically, the following changes have been made:

1. The molecular docking Results section has been revised to clarify that docking scores represent model-dependent computational estimates and cannot be directly interpreted as experimentally measured binding affinities, binding free energies, or inhibitory potencies.

2. The previous terminology of "high-affinity," "moderate-affinity," and "lower-affinity" binders has been replaced with more appropriate descriptions such as "top-ranked compounds based on docking score," "intermediate-ranked compounds," and "compounds with relatively less favorable docking scores." This change was made to avoid implying that the selected docking-score ranges represent experimentally validated affinity thresholds.

3. Statements suggesting that phytochemicals "outperformed" or possessed "stronger binding" than approved EGFR inhibitors have been revised. The manuscript now states that certain phytochemicals received more favorable docking scores than the reference compounds under the same computational docking protocol, while explicitly emphasizing that this numerical difference does not demonstrate superior experimental binding affinity, inhibitory potency, or therapeutic efficacy.

4. The comparison with Osimertinib, Erlotinib, Afatinib, and Lazertinib has been retained as a computational reference comparison rather than as evidence of pharmacological superiority. We now explicitly state that similar or more favorable docking scores indicate only relative ranking within the applied computational framework and do not establish equivalent or superior biological activity.

5. The interpretation of compounds with intermediate and relatively less favorable docking scores has also been revised. We no longer infer pharmacological relevance solely from their docking-score ranges. Instead, we emphasize that docking results should be considered together with binding-mode plausibility, protein–ligand interaction patterns, physicochemical properties, ADMET characteristics, toxicity predictions, and molecular dynamics behavior.

6. A clarification has been added to the docking results and Table 3 indicating that docking scores should be interpreted as relative computational prioritization metrics and not as direct measures of experimental binding affinity or inhibitory potency.

7. The Discussion has been revised to explicitly acknowledge that differences in docking scores cannot establish the superiority of the phytochemicals over clinically used EGFR-targeting drugs. We now emphasize that experimental biochemical assays, including direct EGFR kinase inhibition measurements, are required to determine actual binding affinity and inhibitory potency.

These revisions have substantially moderated the interpretation of the docking results and aligned our conclusions with the limitations of computational docking. We now use the docking analysis primarily as an initial prioritization step within a broader integrated computational workflow rather than as standalone evidence of EGFR inhibitory activity or therapeutic efficacy. We appreciate the reviewer for highlighting this issue, which has helped us improve the scientific rigor and precision of the manuscript.

Comment 3: The use of only the wild-type EGFR structure (PDB: 1M17) does not adequately represent clinically important resistant NSCLC mutations such as L858R, T790M, and C797S, reducing the clinical r

Attachments

Attachment

Submitted filename: Response to the Editor and Reviewers.docx