De Novo Transcriptome Assembly of Cedar (Cedrela odorata L.) and Differential Gene Expression Involved in Herbivore Resistance

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Manuscript entitled “De novo Transcriptome Assembly of Cedar (Cedrela odorata L.) and Differential Gene Expression Involved in Herbivore Resistance” showing the de novo transcriptome assembly results of the timber tree, and find some DEGs. That’s good for cedar research. And I have several comments on this manuscript:

1, Neither in the methods nor in the results sections, I didn’t find information on sequencing data availability. Did the authors deposit raw data of the RNA-seq sequencing? If not, authors should provide raw data in online database such as SRA database in NCBI.

2. Similar with above question, authors provide DEseq2 results in supplementary. However, I could not find gene ID or gene sequence with DEGs, maybe authors could upload Trinity results (.fasta file) as supplementary. And in the S2.upregulated.csv file, the “Transcritos” column should not be deleted.

3. Figure 1. show the most abundant GO terms like “metabolic process” is over 5000 genes. However, in Line 212-215, authors write “the most abundant with 165,75,5, respectively”, please check those numbers.

4. Figure 2. show DEGs in cedar by DEseq2, authors could use heatmap or volcano map instead this style.  Because in Line 237-240, authors said “the 10 transcripts with the lowest expression levels (Figure 2)”, however I can see more than 10 transcripts be marked in the Figure2.

5. Figure 3. show qRCR validation of RNA-seq. Usually authors should show gene expression FPKM/TPM/counts value or heatmap of RNA-seq. So we could see whether those gene exhibit the same expression pattern by RNA-seq and qPCR. 

Author Response

1, Neither in the methods nor in the results sections, I didn’t find information on sequencing data availability. Did the authors deposit raw data of the RNA-seq sequencing? If not, authors should provide raw data in online database such as SRA database in NCBI.

1. Libraries were deposited in NCBI under the BioProject number PRJNA1134724. This information was written down in material and methods, Lines 122-123.

2, Similar with above question, authors provide DEseq2 results in supplementary. However, I could not find gene ID or gene sequence with DEGs, maybe authors could upload Trinity results (.fasta file) as supplementary. And in the S2.upregulated.csv file, the “Transcritos” column should not be deleted.

2. Trinity results and transcriptome were deposited in NCBI under the BioProject number PRJNA1134724.

3, Figure 1. show the most abundant GO terms like “metabolic process” is over 5000 genes. However, in Line 212-215, authors write “the most abundant with 165,75,5, respectively”, please check those numbers.

3. Numbers were check and chart was edited with right information, Figure 1.

4. Figure 2. show DEGs in cedar by DEseq2, authors could use heatmap or volcano map instead this style.  Because in Line 237-240, authors said “the 10 transcripts with the lowest expression levels (Figure 2)”, however I can see more than 10 transcripts be marked in the Figure2.

4. A volcano graphic was added (Figure 2) and text in lines 237-240 were corrected.

5. Figure 3. show qRCR validation of RNA-seq. Usually authors should show gene expression FPKM/TPM/counts value or heatmap of RNA-seq. So we could see whether those gene exhibit the same expression pattern by RNA-seq and qPCR. 

5. Supplementary material (S4) and HeatMap (Figure 2) were included to improve explication of results. If it is necessary, it could be upload to figshare.

Reviewer 2 Report

Comments and Suggestions for Authors

Reviewer comments:

The manuscript entitled “De novo Transcriptome Assembly of Cedar (Cedrela odorata L.) and Differential Gene Expression Involved in Herbivore Resistance”, has been reviewed and found this study reports transcriptome and differential expression of genes involved in herbivore resistance. The manuscript's writing quality and structure could be significantly improved. Furthermore, it lacks a comprehensive range of analyses and makes no recommendations for future solutions or directions based on the results. Improving the clarity and coherence of the writing, broadening the scope of analyses, and including recommendations for future research would significantly strengthen the manuscript.

Manuscript needs to be improved rigorously and based on my analysis I recommend major revision.

Major Revision:

1-     The significance of the findings should be more explicitly stated in the results section of the abstract. It would be better to provide an explanation of the implications of overexpressed and underexpressed genes rather than simply listing figures.

2-     The introduction part of the manuscript lacks coherence and continuity. The paragraphs should be better connected to one another for a more seamless narrative flow.

3-     Carefully observe and italicize scientific names across the manuscript e.g. Chrysobothris yucatanensis. Line#199.

4-     On which basis the control (CK) is different from other healthy plants? Authors are already comparing infested samples with that of healthy samples, then what’s the purpose to compares with another healthy sample as ck? Line# 245-256, figure 3.

5-     The sentence needs to be rewritten in a correct way. Line# 281.

6-     A long yet useless sentence should be rewritten “this seems to contradict other authors”. You don’t need to target or contradict authors, you should discuss the research and studies. Line#290.

7-     Authors should provide the details about KEGG pathways, including which particular pathway was impacted in relation to others, e.g. Phenylpropanoids, glucosinolate pathway.

8-     Provide a figure regarding the KEGG pathway in the main text of the MS.

9-     The paragraph denoting different subpoints i.e. A, B, C should be written in a direct way. Line# 279-288.

1- Although the methodologies are mentioned, including a concise justification for selecting RNAseq and qPCR would improve comprehension.

1- The discussion part requires improved cohesion and logical progression. The transition between different facets of the study, such as the methodologies and results, can be more seamless. For example, the sentence structure can be modified to ensure that each point is logically connected to the preceding one.

1- Supplementary data isn’t arranged in a formal way. Authors should seriously arrange the supplementary data in the form of tables or other settings.

1-  

Comments on the Quality of English Language

The manuscript writing quality lacks coherence and continuity, especially in the introduction and discussion part. For a more cohesive narrative the paragraphs should be better linked, creating a smoother flow of ideas.

Author Response

1,The significance of the findings should be more explicitly stated in the results section of the abstract. It would be better to provide an explanation of the implications of overexpressed and underexpressed genes rather than simply listing figures.

The importance and implication of the findings were included in lines 17-22 of the summary.

“Differential gene expression showed 170 genes with 245 functions: 165, 75, and 5 were molecular functions, biological processes, and cellular components, respectively. Trees produce toxins and volatile compounds to defend against herbivorous, which are modulated by signaling pathways and gene expression, related to molecular functions and biological processes. The small number of genes identified as cellular components suggests minimal changes in cellular structure in response to borer attack”.

2,The introduction part of the manuscript lacks coherence and continuity. The paragraphs should be better connected to one another for a more seamless narrative flow.

We added connectors to the paragraphs in the introduction and changed the information on lines 39-41 with information on cedar and moved the text from reference 10 to lines 84-87. In addition, we added new references on lines 44-46 and 82-84.

3,Carefully observe and italicize scientific names across the manuscript e.g. Chrysobothris yucatanensis. Line#199.

All scientific names were revised and italicized.

4,On which basis the control (CK) is different from other healthy plants? Authors are already comparing infested samples with that of healthy samples, then what’s the purpose to compares with another healthy sample as ck? Line# 245-256, figure 3.

The control is a healthy plant. Validation of gene expression is evaluated by 2-ΔΔCt relative quantification method using qRT-PCR. This method requires both a constitutive gene and a healthy sample to be used as a reference and, to be able to determine the relative expression level with respect to the rest of the samples, treatments or tissues, as in this case. When there are only two conditions, healthy and infested, a healthy sample should be used as a control.

5,The sentence needs to be rewritten in a correct way. Line# 281.

Text was rewritten: “In adittion, upregulated gene families identified in this study can be grouped according to their main function. The first category is related to synthesis of defensive secondary compounds, including terpene synthase, cytochrome P450, and possible lysine decarboxylase. The second category is associated with signal transduction and regulation of gene expression, including NAD-dependent epimerase/dehydratase, protein kinase domain, zinc-binding dehydrogenase, and response regulator receiver domain. The third category is related to oxidation and reduction of compounds, including zinc-binding dehydrogenase and GroEs-like alcohol dehydrogenase domain. These protein families are directly or indirectly involved in the plant's response to herbivore attack.”

6,A long yet useless sentence should be rewritten “this seems to contradict other authors”. You don’t need to target or contradict authors, you should discuss the research and studies. Line#290.

Sentence was change to “At first glance, our findings suggest that these genes could have a minimal participation in the plant's response to herbivore attack”.

7,Authors should provide the details about KEGG pathways, including which particular pathway was impacted in relation to others, e.g. Phenylpropanoids, glucosinolate pathway.

We appreciate the observations, however, it was not considered that  KEGG pathways were included because it is not common to include this information in differential gene expression papers, the focus of study was on genes related to the attack of herbivorous insects and perhaps the objective could be misunderstood if this information is included. All in order for the journal's readers to understand the meaning of the work.

8,Provide a figure regarding the KEGG pathway in the main text of the MS.

Related to answer No. 7.

9,The paragraph denoting different subpoints i.e. A, B, C should be written in a direct way. Line# 279-288.

Paragraph was modified by “The identified downregulated gene families can be grouped according to their main function. The first category is related to synthesis of defensive secondary compounds, including terpene synthase, cytochrome P450, and possible lysine decarboxylase. The second category is associated with signal transduction and regulation of gene expression, including NAD-dependent epimerase/dehydratase, protein kinase domain, zinc-binding dehydrogenase, and response regulator receiver domain. The third category is related to oxidation and reduction of compounds, including zinc-binding dehydrogenase and GroEs-like alcohol dehydrogenase domain. These protein families are directly or indirectly involved in the plant's response to herbivore attack.”

1,Although the methodologies are mentioned, including a concise justification for selecting RNAseq and qPCR would improve comprehension.

RNAseq enables global gene expression when genomic information is not available and it can obtain complete coverage of transcripts, structure of exons and alternative splicing. qPCR is a methodology characterized by high specificity and sensitivity with many applications. The validation of differential expression obtained by transcriptome analysis is one of these applications, and is also widely used in reports of this type of work. It was not considered to include a justification of the methodology since it is currently accepted by the scientific community and it is not common to include a clarification in the manuscripts.

1,The discussion part requires improved cohesion and logical progression. The transition between different facets of the study, such as the methodologies and results, can be more seamless. For example, the sentence structure can be modified to ensure that each point is logically connected to the preceding one.

The reviewer's recommendation was addressed to improve cohesion and logical progression, mainly in the lines: 262-263; 265-266; 278-279; 329-230; 342-343.

1,Supplementary data isn’t arranged in a formal way. Authors should seriously arrange the supplementary data in the form of tables or other settings.

Supplementary data was arranged in a formal way with new tables in Microsoft Excel. Supplementary materials 1-4.

The manuscript writing quality lacks coherence and continuity, especially in the introduction and discussion part. For a more cohesive narrative the paragraphs should be better linked, creating a smoother flow of ideas.

The reviewer's recommendation was addressed to improve cohesion and logical progression, mainly in the lines: 262-263; 265-266; 278-279; 329-230; 342-343.

English was edited.

Reviewer 3 Report

Comments and Suggestions for Authors

Comments to the Authors

In this study, the authors reported reported de novo assembly of cedar (Cedrela 14

odorata L.) transcriptome and differential expression of genes involved in herbivore resistance. A total of 325.6 million reads were obtained, and 127,031 (97.47%) sequences were assembled. Among 1,614 genes reported in embryophytes, 1,404 (87%) complete genes were detected. In the identification of 220 differentially expressed genes, 161 showed underexpression and 59 showed overexpression. This is the first report of Cedrela odorata L. transcriptome focusing on genes overexpressed in healthy plants and underexpressed in infested plants that may be a better strategy for identifying basal genes related to herbivore resistanse. However, there some major revisions should be clarified or made in the manuscript, as following:

1. Line89，”Plant material was obtained from five healthy cedar plants and six cedar plants showing”, authors should describe the growth status of trees. Because overexpress or underexpress of the genes in the plants may be associated with the growth status of trees.

2. Line 159-163, “allowing a focus on transcriptional variations between these genes in healthy and C. yucatanensis-infested plants.”, “enabling better identification of genes related to herbivore resistance”. How can the author exclude the influence of these factors inherent to phenological stages, environmental conditions, and management practices? The growth status and living conditions of plant samples should be clarified clearly in the manuscript

Author Response

1, Line89，”Plant material was obtained from five healthy cedar plants and six cedar plants showing”, authors should describe the growth status of trees. Because overexpress or underexpress of the genes in the plants may be associated with the growth status of trees

Information of the growth status of plants is completed. “Plant material was obtained from five healthy cedar plants and six cedar plants showing signs of pest attack by Chrysobothris yucatanensis from the San Felipe Bacalar nursery of the Chetumal-INIFAP experimental field in Chetumal, Quintana Roo, Mexico. All of plants were two years of growth, planted in individual plastic pots, placed under mesh-shade and they had equal management practices”.

2. Line 159-163, “allowing a focus on transcriptional variations between these genes in healthy and C. yucatanensis-infested plants.”, “enabling better identification of genes related to herbivore resistance”. How can the author exclude the influence of these factors inherent to phenological stages, environmental conditions, and management practices? The growth status and living conditions of plant samples should be clarified clearly in the manuscript.

All plants had equal conditions. The materials section was completed with information on conditions, management and phenological stages of the plants to clarify this point, lines 92-94. “All of plants were two years of growth, planted in individual plastic pots, placed under mesh-shade and they had equal management practices”.

Reviewer 4 Report

Comments and Suggestions for Authors

In “De novo Transcriptome Assembly of Cedar (Cedrela odorata L.) and Differential Gene Expression Involved in Herbivore Resistance”, the authors perform transcriptome sequencing analysis on cedar plants. They compare transcriptomic expression levels of herbivore-related genes between five healthy plants and six that are infested with Chrysobothris woodboring beetles. Then they validate a few of the overexpressed genes using qRT-PCR. There were differences in expression in protein families related to recognition, signaling and transcription: chitin recognition, JAZ motifs, and response regulator receivers. There were also families related to defense: allene oxide cyclase, lipoxygenases, peroxidases. There were also families related to growth, development and metabolism: Terpenoid proteins and cytochrome P450. However, when they validated expression patterns with qPCR, they found that TERS and LIP01 were overexpressed in all samples, not just attacked samples. This is a novel study with results that could be important in regulating and addressing bark beetle problems and developing resistant strains of cedar, however it needs improvement in at least three areas:

 

1)      In the methods, the authors describe their lower limit for fold change level at 2 and a false discovery rate of 0.05 (not really relevant for DESeq2 results—should use adjusted p value). However, in the accompanying supplementary tables, there are many proteins with log fold change levels <2 and p adjusted values >0.05. If the authors would like to use log fold change differences >1, they should say so. There are also several times throughout the manuscript where authors refer to the supplementary tables and the numbers in the manuscript don’t match the numbers in the files. It is also unclear when the authors describe over and under-expression whether they are talking about upregulation in healthy or attacked samples. Please clarify. It also seems like S2 might be the wrong file? It doesn’t have any gene families listed, and is a .tsv file that is listed as a .csv. I think it would be helpful to take maybe the top 5 up and downregulated gene families and put them in a in-text table with their GO terms and WEGO if they are the ones you are talking about the most.

2)      Unclear methods around qPCR.. The methods in line 187 suggest that only one healthy plant was used as a control sample, yet the legend from Figure 3 suggests there are four healthy samples and two infested samples compared. What is the difference between the healthy samples and the reference? How were the genes selected for validation? Are they related to the overexpressed ones in S1? It also seems like the validation did not confirm overexpression of TERS and LIP01. Am I understanding correctly? Please add to discussion if this was unexpected? Since all samples were from the same nursery is it possible that they signaled to each other and are all responding to attack? Axes on Figure 3 too small to read

3)      Introduction: The authors list several proteins that other researchers have found as significant. This is great, but it would be better if also in the discussion, they come back and compare how their results compare to the other research findings. Discussion should also be better organized—talk about what gene families you expected to see and possible explanation for why not.

 

 

Additional fixes that would improve the manuscript:

·         Line 200: For sequencing results, list mean, min and max for base pairs and sequences assembled.

·         Line 351: Unclear what better strategy means

See additional comments in attached pdf.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

A few grammatical errors in the abstract/introduction.

Author Response

1)      In the methods, the authors describe their lower limit for fold change level at 2 and a false discovery rate of 0.05 (not really relevant for DESeq2 results—should use adjusted p value). However, in the accompanying supplementary tables, there are many proteins with log fold change levels <2 and p adjusted values >0.05. If the authors would like to use log fold change differences >1, they should say so. There are also several times throughout the manuscript where authors refer to the supplementary tables and the numbers in the manuscript don’t match the numbers in the files. It is also unclear when the authors describe over and under-expression whether they are talking about upregulation in healthy or attacked samples. Please clarify. It also seems like S2 might be the wrong file? It doesn’t have any gene families listed, and is a .tsv file that is listed as a .csv. I think it would be helpful to take maybe the top 5 up and downregulated gene families and put them in a in-text table with their GO terms and WEGO if they are the ones you are talking about the most.

The logFoldChange value was corrected to 1 and p-adj < 0.05. Simultaneously, data were reorganized into a more accessible format using an Excel spreadsheet. These adjustments to LFC and p-adj uncovered errors in the data analysis, revealing that the interpretation of over- and under-expressed genes had been inaccurately assessed. We appreciate your observation as it enabled us to rectify a critical error in the results and discussion of this article.

2,Unclear methods around qPCR.. The methods in line 187 suggest that only one healthy plant was used as a control sample, yet the legend from Figure 3 suggests there are four healthy samples and two infested samples compared. What is the difference between the healthy samples and the reference? How were the genes selected for validation? Are they related to the overexpressed ones in S1? It also seems like the validation did not confirm overexpression of TERS and LIP01. Am I understanding correctly? Please add to discussion if this was unexpected? Since all samples were from the same nursery is it possible that they signaled to each other and are all responding to attack? Axes on Figure 3 too small to read.

The control is a healthy plant. Validation of gene expression is evaluated by 2-ΔΔCt relative quantification method using qRT-PCR. This method requires both a constitutive gene and a healthy sample to be used as a reference and, to be able to determine the relative expression level with respect to the rest of the samples, treatments or tissues, as in this case. When there are only two conditions, healthy and infested, a healthy sample should be used as a control.

Plants were planted in individual plastic pots and placed under mesh-shade”.

3. Introduction: The authors list several proteins that other researchers have found as significant. This is great, but it would be better if also in the discussion, they come back and compare how their results compare to the other research findings. Discussion should also be better organized—talk about what gene families you expected to see and possible explanation for why not.

We greatly appreciate your comments and observations. In the realm of forest trees, such studies are indeed scarce. Most of the data against which we compared these gene families come from studies conducted on perennial plants, where the protein family descriptions were established through biochemical methods. In the case of Cedar, there is no well-annotated transcriptome available for comparing the genes detected by us against those reported in databases. Furthermore, phylogenetically close species belong to genera that are phylogenetically and geographically distant, complicating further comparisons with the limited information available in the literature. Hence, the gene search was conducted via homology with all proteins available in Uniprot.

 4, Line 200: For sequencing results, list mean, min and max for base pairs and sequences assembled.

These values were added in lines 216-218 "The median conting length was 532 pb, the average conting length was calculated at 942.63 pb and N50 was 1,623 pb."

5, Line 351: Unclear what better strategy means

Sentece was completed to improve understanding “Focusing on genes overexpressed in healthy plants and underexpressed in infest-ed plants might be a better strategy for identifying basal genes related to herbivore defense than traditional studios that analyze gene overexpressed in infested plants”

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thanks for the author's response to my 1st round questions/suggestions.I still have several questions/suggestions:

1,   I saw authors upload RNA-seq raw data in NCBI under the BioProject number PRJNA1134724 on 11-Jul-2024. It's good for reader. However, there are only 10 library data can be found in the website, while 11 library be used in this paper. Please check the the BioProject data.

2,  Authors said Trinity results and transcriptome were deposited in NCBI under the BioProject number PRJNA1134724.  Please give the link to the Trinity results (.fasta file). 

3, I said "  Figure 1. show the most abundant GO terms like “metabolic process” is over 5000 genes. However, in Line 212-215, authors write “the most abundant with 165,75,5, respectively”, please check those numbers.   "   And authors said "3. Numbers were check and chart was edited with right information, Figure 1."        However,  I still see more than 5000 genes under GO term ''binding", that number is far more diffrent  with what author write in the text. DESeq2 detected 220 genes as DEGs, so in the Figure 1, gene numbers should less than 220.  I guess authors used the whole Trinity genes to do GO analysis instead of DEGs. 

4. Authors added volcano graphic in (Figure 2) . That's a good job. However, authors should deleted the old graph with "baseMean, log2FoldChange", the old graph have be used in another online paper "Optimized method for differential gene expression analysis in non-model species"( https://pubmed.ncbi.nlm.nih.gov/37920871/ ) in Fig.1B. 

 

Author Response

1,   I saw authors upload RNA-seq raw data in NCBI under the BioProject number PRJNA1134724 on 11-Jul-2024. It's good for reader. However, there are only 10 library data can be found in the website, while 11 library be used in this paper. Please check the the BioProject data.

All 11 libraries were submitted to the National Center for Biotechnology Information (NCBI) for inclusion in their database. With regard to the H11 library, an error was identified in the processing conducted by NCBI. To remove the library and submit it anew, the following SRA protocol was carried out. Consequently, it falls upon NCBI to ensure the library's continued availability.

2,  Authors said Trinity results and transcriptome were deposited in NCBI under the BioProject number PRJNA1134724.  Please give the link to the Trinity results (.fasta file). 

NCBI is processing the cedar transcriptome, but you can find Trinity results in figshare, available in: https://doi.org/10.6084/m9.figshare.26352232. This information was added in line 128.

3, I said "  Figure 1. show the most abundant GO terms like “metabolic process” is over 5000 genes. However, in Line 212-215, authors write “the most abundant with 165,75,5, respectively”, please check those numbers.   "   And authors said "3. Numbers were check and chart was edited with right information, Figure 1."        However,  I still see more than 5000 genes under GO term ''binding", that number is far more diffrent  with what author write in the text. DESeq2 detected 220 genes as DEGs, so in the Figure 1, gene numbers should less than 220.  I guess authors used the whole Trinity genes to do GO analysis instead of DEGs. 

The figure 1 were changed in GO terms.  

4, Authors added volcano graphic in (Figure 2) . That's a good job. However, authors should deleted the old graph with "baseMean, log2FoldChange", the old graph have be used in another online paper "Optimized method for differential gene expression analysis in non-model species"( https://pubmed.ncbi.nlm.nih.gov/37920871/ ) in Fig.1B. 

Figure 1B differs from the previous graph that was utilized in another online paper. It is possible you were confused by the apparent similarity of the two graphs, but they are, in fact, not the same.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have made substantial revisions in response to the reviewers' suggestions. This work adds a significant contribution to the field of transcriptome and differential expression of genes involved in herbivore resistance in Cedar. The study is anticipated to enhance practical applications and better strategy for identifying basal genes related to herbivore defense than traditional studios that analyze genes overexpressed in infested plants.

Overall, the manuscript is now of good quality and satisfies the requirements for publication in CIMB. I suggest acceptance of the manuscript with no further revisions required.

Comments on the Quality of English Language

Minor editing of English language required

Author Response

Thank you for your comments 

Editing of English language was carried out.

Reviewer 3 Report

Comments and Suggestions for Authors

Well revised

Author Response

Thank you for your comments 

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

Please make sure all 11 library raw data could be download by reader. 

Author Response

1, Please make sure all 11 library raw data could be download by reader.

Now, all of 11 library raw data can be download.

Thank you for your all comments.

Round 4

Reviewer 1 Report

Comments and Suggestions for Authors

No more question. Please double check the  Fig 1.  with gene numbers.

Author Response

No more question. Please double check the  Fig 1.  with gene numbers.

Dear reviewer, we appreciate your comments on Fig, 1.

We have corroborated the number of genes present and have not made any changes to the figure. The reason why the total number of genes plotted apparently exceeds the total number of genes specified in the text is due to how GO terms work, since each gene can have more than one function and in turn within each biological, cellular or molecular function can have more than one term. This is clearly illustrated with an example provided by WEGO (https://wego.genomics.cn/), if you plot the demonstration data you can see that if we take the values of Demo1; 1,679 genes were annotated, but the total sum of functions exceeds 3,158. Now, if you look at the plotted values, the sum of all the bars far exceeds 1,679 annotated genes. This becomes even clearer when, if we take as an example the cellular components of Demo1, for this particular function 532 functions are reported. However, when we look at the graph in detail and if we add approximately the values of the number of genes in the bars for the terms cell, part of the cell and membrane (terms of the cellular functions) there are more than 1,000 genes. This is the reason why the values in Fig. 1 supposedly do not seem to reflect what is reported. Hopefully this clarification will help to dispel your doubts. Best regards. 

 

Round 5

Reviewer 1 Report

Comments and Suggestions for Authors

No more question.