Genetic Optimization in Uncovering Biologically Meaningful Gene Biomarkers for Glioblastoma Subtypes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Comments and Suggestions for the manuscript 2850548:

In this manuscript, the authors investigated gene biomarkers for glioblastoma (GBM) subtypes, including proneural and mesenchymal subtypes, by using a genetic optimization algorithm. This algorithm integrates differential gene expression and gene variability, aiming to select genes that differentiate between these GBM subtypes. The method uses single-cell RNA-sequencing (scRNA-seq) data and aims to uncover significant genes related to GBM.

Suggestions:

Methodology

1)      Please add the workflow in the Methodology section since the authors have used several algorithms to perform this project, so adding the scheming picture is easy for readers to understand the working pathways.

2)      Most importantly, the authors have to list the software, pipelines, datasets used in this study, including version number or URL. This is a standard for all publication.

Results

3)      The authors should briefly describe how you obtained Figure 1 and Figure 4 and the characteristic of the results or picture shown.

4)      The authors should provide some details of methods to obtain DEGs, HVGs, fitness scores, and a panel of 92 genes, including software and datasets.

5)      Please provide details ON how you finally selected six key genes as GBM markers for PN and MES subclasses.

Discussion

The authors should thoroughly discuss the findings and novelty of this study. Since 2008 GBM PN and MES subtypes have been studies for a long time, including scRNAseq analysis, and the authors cannot help but highlight your findings and implications.

Author Response

Comment 1:

The reviewer suggested adding a workflow diagram in the Methodology section to illustrate the various algorithms used throughout the project, making it easier for readers to understand the working pathways.

Answer:

Thank you for your constructive suggestion to add a workflow diagram in the Methodology section. We have now included a detailed workflow diagram that illustrates the sequence of algorithms and methodologies used in our study. This addition aims to enhance readers' understanding of our approach in identifying gene biomarkers for glioblastoma subtypes. We believe this improvement significantly clarifies the research process outlined in our manuscript.

 

Comment 2:

Most importantly, the authors have to list the software, pipelines, datasets used in this study, including version number or URL. This is a standard for all publication.

Answer:

We agree with the reviewer recommendation to detail the software, pipelines, and datasets used in our study, including version numbers or URLs, recognizing this as a standard practice for ensuring reproducibility and transparency in research publications. Accordingly, we have compiled all relevant details, including version numbers and access links, on a dedicated GitHub repository. This repository provides comprehensive information on the tools and data utilized in our research, facilitating further exploration and verification of our work by the research community. The repository is accessible at the following URL: PaplomatasP/GeneSelector: This repository hosts a Genetic Algorithm-based tool for analyzing scRNA-seq data, pinpointing DEGs and HVGs to uncover genes with significant variance and biological importance. (github.com)

Besides the link is added in paper too.

 

Comment 3:

The authors should briefly describe how you obtained Figure 1 and Figure 4 and the characteristic of the results or picture shown.

Answer:

We have updated the manuscript to include a brief description of how we obtained Figure 1 and Figure 4, along with the characteristics of the results or pictures shown.

 

Comment 4:

The authors should provide some details of methods to obtain DEGs, HVGs, fitness scores, and a panel of 92 genes, including software and datasets.

Answer:

Thank you for your request for additional details on our methodologies concerning the identification of DEGs (Differentially Expressed Genes), HVGs (Highly Variable Genes), and the computation of fitness scores. We are providing a more detailed explanation below:

DEGs: For the identification of DEGs between glioblastoma subtypes, we utilized the Wilcoxon Rank Sum test, as implemented in the Seurat package. This non-parametric test is widely accepted for its ability to identify genes with statistically significant differences in expression levels across subtypes, thereby highlighting potential biomarkers for further investigation. Our selection criteria for DEGs included a stringent adjustment for multiple comparisons to control the false discovery rate (FDR), ensuring that identified genes are highly likely to be truly differentially expressed. For a detailed understanding of how the Wilcoxon test is applied in our context using the Seurat package, please refer to the Seurat documentation available at Differential expression testing (satijalab.org)

HVGs: The identification of HVGs was performed through variance analysis within our genetic algorithm. This step is crucial for capturing genes that exhibit significant expression variability across samples, indicative of their potential regulatory roles in glioblastoma subtypes. Our approach for HVG selection was based on the dispersion of expression values, focusing on genes that not only show high variance but also maintain biological relevance in the context of glioblastoma.

Fitness Scores: The computation of fitness scores within our genetic algorithm integrates key metrics such as variance, p-value, and log fold change (logFC), with a weighted scoring system that assigns 50% to variance, 30% to logFC, and 20% to the p-value score. This weighting reflects the relative importance of each metric in determining gene significance, balancing statistical evidence with biological variability and effect size. Scores undergo normalization to ensure comparability, with an average of these normalized, weighted scores computed to identify genes that are statistically significant and biologically meaningful.

Panel of 92 Genes: In response to the request for details regarding the panel of 92 genes, we have uploaded a CSV file, “Res_92Genes.csv”, to our GitHub repository. This file not only lists the 92 genes identified through our analysis but also includes detailed annotations on their biological functions and potential relevance to glioblastoma. Furthermore, we provide accompanying scripts and datasets, ensuring transparency and reproducibility of our analysis. It is readily available for review and further analysis at GeneSelector/Res_92Genes.csv at Master · PaplomatasP/GeneSelector (github.com)  Besides the link is added in paper too.

 

 

Comment 4

Please provide details ON how you finally selected six key genes as GBM markers for PN and MES subclasses.

Answer:

We have updated the manuscript to include details on how we ultimately selected the six key genes as GBM markers for the PN and MES subclasses.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Title: Genetic optimization in uncovering biologically meaningful gene biomarkers for glioblastoma subtypes

 

Summary: Author has performed an interesting research study depicting a genetic optimization algorithm for the identification of key gene that differentiate between proneural and mesenchymal subset of glioblastoma multiforme. Study is not only scientifically sound but also discussion rich. However, there are some suggestions provided to the author for the better understanding of the study.

 

 

1)        Author is advised to prepare an algorithmic flowchart depicting the stepwise events in methodology and results section for the better understanding of the study flow

2)        Author has described the possible role of higher mRNA expression of VAMP5, CLIC1, PLXND1, MSN, METRNL and LITAF genes in mesenchymal gene of glioblastoma multiforme. Author is suggested to demonstrate these findings with the mean of illustrative figure.

3)        Author is suggested to depict figure number 6 in its full length so that it can better be understood

Author Response

Comment 1:

Author is advised to prepare an algorithmic flowchart depicting the stepwise events in methodology and results section for the better understanding of the study flow.

Answer:

We appreciate the valuable suggestion to include an algorithmic flowchart, which aligns with the recommendation made by Reviewer #1. Accordingly, we've added a concise workflow diagram in our manuscript, illustrating our analytical process for identifying glioblastoma gene biomarkers. This aims to improve the clarity and understanding of our study's methodology.

 

Comment 2:

Author has described the possible role of higher mRNA expression of VAMP5, CLIC1, PLXND1, MSN, METRNL and LITAF genes in mesenchymal gene of glioblastoma multiforme. Author is suggested to demonstrate these findings with the mean of illustrative figure.

Answer:

We appreciate your insightful comment regarding the manuscript. As per your suggestion, we have revised the manuscript to include additional description and discussion.

 

Comment 3:

Author is suggested to depict figure number 6 in its full length so that it can better be understood

Answer:

The image contains a thorough analysis, and zooming in reveals more details. However, enlarging it further risks losing critical information due to the complexity of the analysis.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The authors Paplomatas et al., discussed a very crucial genetic optimization methods to identify novel biomarkers in Glioblastoma. Overall manuscript is very strong with appropriate literature addressed.

There are a vey few minor corrections needed:

1. In Page 7, Line 267, we see a commented reference in there which was already added in the references. I think it was an honest mistake in editing. 

2. The figures are little blurred in Fig4, 5 and 6 . This makes it harder to view and interpret. It needs to be corrected.

The rest is very innovative and interesting to explore the biomarkers in GBM. 

Author Response

Comments 1:

  In Page 7, Line 267, we see a commented reference in there which was already added in the references. I think it was an honest mistake in editing.

Answer:

Thank you for bringing this to our attention. Indeed, it was an oversight, and we appreciate your thorough review. The commented reference on Page 7, Line 267, has been removed to avoid redundancy.

 

Comments 2:

The figures are little blurred in Fig4, 5 and 6 . This makes it harder to view and interpret. It needs to be corrected.

Answer:

We acknowledge your concern regarding the clarity of the figures in Fig 4, 5, and 6. We apologize for any inconvenience this may have caused. To ensure better visibility and 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

In this version of the manuscript, some details about methods and resources are provided which will be helpful to the reader.

There may be some errors in the version received, for example the figures for Figures 1 and 2 and their legends have been removed without any new replacements. If there is no new one, keep the old one, otherwise, the manuscript will not be completed and will not match the context. Please check these carefully.

Comments on the Quality of English Language

English language is good enough for this publication.

Author Response

Thank you for your valuable feedback. There may have been an oversight on our part, for which we apologize. We are now submitting the final and correct version of our document. We appreciate your understanding and patience.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Author has satisfactorily resolved the queries 

Author Response

We have made an update to our manuscript and conducted thorough proofreading to address your concerns. We also want to express our gratitude for your positive comments and constructive criticism. Thank you for helping us improve the quality of our work.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you for addressing the comments.  The authors have thoroughly checked and answered all the comments. The signaling pathway is still blurred. 

Author Response

We have made efforts to improve the resolution of the KEGG pathway as you recommended. However, we acknowledge there might still be room for enhancement in terms of quality. We invite you to review the updated pathway visualization at our GitHub repository, which may offer a better resolution. You can view it at the following link: GeneSelector/Plots/hsa05200..png at Master · PaplomatasP/GeneSelector (github.com). We appreciate your feedback and are committed to further refining our work.

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

Everything is fine. Thanks for your consideration.