Seed Protein and Oil QTL in a Prominent Glycine max Genetic Pedigree: Enhancing Stability for Marker Assisted Selection

Round 1

Reviewer 1 Report

Research on soybean is important because it is an excellent source of plant protein. The publication submitted for review tries to enrich knowledge on the identification and confirmation of seed protein and seed oil QTLs. After reading such a publication, I expect that it will provide new information about the biology of crop plants and it is expected to improve capacities for modelling plant adaptation responses to better predict changes in plant performance and crop improvement and management strategies. 

 

Although the publication is well structured, with many tables, I notice a lack of basic information. My remarks:

1.     There is a lack of clear, strong hypothesis – please complete.

2.     I don't understand the methodology:

Line 34: Over the two year (there is an error here, it should be “two years of study”) study which included a single location the first year and three locaions the second year, a total of 27 QTL for protein and oil were detected.

 

What is the justification for such an experiment pattern, why there were not three locations in both cases? Could this affect the results?

 

3.     How did the genotypes Essex and Williams 82 selected for crossing differ in terms of seed protein and seed oil?

 

4.     A great difficulty in interpreting the results is the lack of information on weather conditions. What was the rainfall and temperature during the experiment?

 

5.     There are editing errors in the manuscript. Table titles should not end with a period, nor should table captions.

 

6.     The Discussion section is very short and basically nothing revealing. Rather, such studies should be part of a larger study, including, for example, a broader proteomic approach or broader phenotypic characterization. In the current arrangement, it is only a report in a narrow scope, not suitable for publication nowadays. Currently, publications are multidisciplinary and bring innovative information to molecular breeding. I suggest expanding your research by including new parameters, maybe (just as an example from me) analysis of physiological traits or others.

 

In conclusion, the manuscript cannot be published in its current form. After the publication will be enriched with additional parameters and the methodological section will be supplemented, the publication may be re-reviewed.

Author Response

Reviewer 1:

 

Research on soybean is important because it is an excellent source of plant protein. The publication submitted for review tries to enrich knowledge on the identification and confirmation of seed protein and seed oil QTLs. After reading such a publication, I expect that it will provide new information about the biology of crop plants and it is expected to improve capacities for modelling plant adaptation responses to better predict changes in plant performance and crop improvement and management strategies. 

 

Although the publication is well structured, with many tables, I notice a lack of basic information. My remarks: Thank you, I hope responses will help to point out the valuable information gained in the study.

1. There is a lack of clear, strong hypothesis – please complete. Given the qualitative nature of the quality    traits being evaluated, the species is best understood when we locate as many genetic markers associated with the quality trait as are possible to build a picture and help to achieve the goal of identify genomic regions associate with these traits. While any study can be more comprehensive, we effectively achieved the objective of expanding knowledge of genomic regions by verifying and confirming critical QTL in a important lineage to US soybean.
2. I don't understand the methodology:

Line 34: Over the two year (there is an error here, it should be “two years of study”) study which included a single location the first year and three locations the second year, a total of 27 QTL for protein and oil were detected.  Thank you. Perhaps a language clarification was needed and I have responded.  Please see lines 35 has been clarified. The was data used from a previous year and one location. The actual study is one year with three locations.

 

What is the justification for such an experiment pattern, why there were not three locations in both cases? Could this affect the results? The study is a one-year study in three locations to detect QTL for protein and oil. The prior year study is referred to show evidence of the confirmed QTL. I have clarified this point in the manuscript and thank you for bringing it to my attention.

 

3. How did the genotypes Essex and Williams 82 selected for crossing differ in terms of seed protein and seed oil? Please refer to Tables 1 and 2 which provides the requested information and clarifies the benefits to such studies in identifying promising lines for genetic improvement. There is a mild difference, Essex is the high parent for protein and Williams 82 is the high parent for oil.

 

4. A great difficulty in interpreting the results is the lack of information on weather conditions. What was the rainfall and temperature during the experiment? The locations and conditions at each research station are described in lines 179-186 of the original manuscript. The rainfall and temperature were not atypical, additional information was not included in the scope of this work because we are not breeding from drought resistance etc.
5. There are editing errors in the manuscript. Table titles should not end with a period, nor should table captions. Thank you for pointing these out, the errors have been addressed. Regarding tables, respectfully I believe the rule is journal specific. Agronomy does allow periods in the table. I am glad to adjust if it will not be flagged as an editorial issue with this particular journal.
6. The Discussion section is very short and basically nothing revealing. Rather, such studies should be part of a larger study, including, for example, a broader proteomic approach or broader phenotypic characterization. In the current arrangement, it is only a report in a narrow scope, not suitable for publication nowadays. Currently, publications are multidisciplinary and bring innovative information to molecular breeding. I suggest expanding your research by including new parameters, maybe (just as an example from me) analysis of physiological traits or others. Thank you for the comment and the opportunity to explain, I have expanded the discussion to tell a better story of why these markers are important.

The focus of this work is an important population foundational for US soybean lineage, thus the narrow focus to gather specific information. The illumination of QTL associated with protein and oil varieties that are in the lineage of many soybean lines in the US is a valid and important pursuit.  Given that we have not cracked the code to effectively improve protein will maintaining oil, we continue to gather and exploit information that could help toward the goal. Successive studies will be built from the knowledge gained in this study and we believe we effectively accomplished the objectives that will aid the soybean community in understanding how we may enhance protein and oil. Every contribution helps us come closer to understanding how to manipulate the trait for production benefit.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

The article presents the results of a QTL mapping study in a biparental soybean population for seed protein and oil content.
The genetic resolution is good, the penotyping intensity is ok and the statistics seem to be solid. The topic fits the scope of MDPI agronomy.
There are several issues which must be adressed before publication (see below).

L53 The correlation between protein and oil content should also be mentioned in this context.
L57 Please adhere to the MDPI style guide when making citations. Please check all instances.
L61 The part about previous QTL mapping studies is a bit lengthy for the introduction. I suggest to only mention the studies involving Essex in this section and move other parts to the discussion or just remove it.
L78-86 This section interrupts the results from previous studies and should be moved.
L99/302 You have only one population and three objectives, which all involve QTL confirmation. This is redundant and confusing. Please simplify.
L114-143 Please shorten the description about population development and propagation. This section should answer where and when the actual trials were conducted exactly but it doesn't.
L133 Line with reference 19 must be mentioned.
L172 Please check the formula:
    - redundant multiplication sign before plus sign,
    - index for B is i,
    - index for T is j,
    - and the error term (with index ij) is missing.
L177 This gives the impression that means across all environments were used for QTL analysis. But in the results section I can find only results from seperate environments, which is ok.
L208 Table 1 should be mentioned before Table 2 in the text.
L220 It is h2, not R2.
L227 Seed weight should be either covered in all sections or removed from this table.
L237 "Based"
L266 Please merge Tables 5 and 6. If you list both phenotypes in Table 4 you can do the same with the results from 2013. A single table with all QTL results would be another solution.
L282 You can draw this conclusion because of high genotypic variance and low GxE variance. Environmental variance per se does not lower heritability.
L299 Please clarify which result exactly allows you to draw this conclusion (transgressive segregation or high h2?). Is it only phenotypic selection?
301-376 It is unnecessary to use so many levels of confirmation for a simple QTL study (verified, confirmed, positionally confirmed + abbreviations). Important is only confirmation within the study (across environments) and confirmation across studies. QTL confirmed only within this study (i.e. Gm 9) get too little attention.
L315 I agree that confirmation across environments is the crucial criterion. But now the criterion became "both years". Does not make sense to me.
L324 Please check citation style.
L342 I don't agree. If you confirm a QTL in other donors you can deduce that the alleles of this QTL may not be too rare and that it functions in different genetic backgrounds.
L383 "pleiotropy"

Author Response

Reviewer 2- Soybean Protein Article

The article presents the results of a QTL mapping study in a biparental soybean population for seed protein and oil content.
The genetic resolution is good, the phenotyping intensity is ok and the statistics seem to be solid. The topic fits the scope of MDPI agronomy.
There are several issues which must be addressed before publication (see below).

L53 The correlation between protein and oil content should also be mentioned in this context.

Yes, the information about the  inverse relationship between protein and oil among some varieties has been added.

L57 Please adhere to the MDPI style guide when making citations. Please check all instances.
Changed to [4-9] instead of [4,5,6,7,8,9].

L61 The part about previous QTL mapping studies is a bit lengthy for the introduction. I suggest to only mention the studies involving Essex in this section and move other parts to the discussion or just remove it. I believe the studies are worthwhile mentioning because they lay the foundation for our research approach.

L78-86 This section interrupts the results from previous studies and should be moved. Thank you for your point, it had been restructured and some deleted.

L99/302 You have only one population and three objectives, which all involve QTL confirmation. This is redundant and confusing. Please simplify. The objectives have been simplified, thank you for pointing out the redundancy.

Changed to: 1) identify new QTL, verify and/or confirm QTL for seed protein and seed oil in the ‘Essex x Williams 82′, 50,000 SNP RIL population (ExW82-50K); 2) test the stability of seed protein and seed oil QTL across multiple environments.

L114-143 Please shorten the description about population development and propagation. This section should answer where and when the actual trials were conducted exactly but it doesn't.

I do understand your view.  The detail for the development of the breeding population were meticulously provided due to the historical importance and as reference which breeders often like to know. Both Essex and Williams 82 are part of the lineage of numerous important varieties in the US. Regarding the actual trial information, please see lines 130-143 describing timing, location and conditions of the research plots for the trials.
L133 Line with reference 19 must be mentioned. Osage added… Thank you
L172 Please check the formula:
    - redundant multiplication sign before plus sign,
    - index for B is i,
    - index for T is j,
    - and the error term (with index ij) is missing. Are you saying you would prefer: Y_ij = µ+B_i+T_j + B*T_{ij +} ε_i(j),

 it has been adjusted.

L177 This gives the impression that means across all environments were used for QTL analysis. But in the results section, I can find only results from separate environments, which is ok. I hope it is more clear now.
L208 Table 1 should be mentioned before Table 2 in the text. Thank you, the correction has been made and some tables reordered.
L220 It is h2, not R2…. Good eyes thanks
L227 Seed weight should be either covered in all sections or removed from this table It has been removed.
L237 "Based" typo corrected
L266 Please merge Tables 5 and 6. If you list both phenotypes in Table 4 you can do the same with the results from 2013. A single table with all QTL results would be another solution. Thank you. Completed for simplification and show consistency… fresh eyes really make a difference to the publication.
L282 You can draw this conclusion because of high genotypic variance and low GxE variance. Environmental variance per se does not lower heritability. I agree and clarified the language, I meant in addition the high heritability would make them more useful. Thanks.
L299 Please clarify which result exactly allows you to draw this conclusion (transgressive segregation or high h2?). Is it only phenotypic selection? Yes, it is phenotypic selection and the identification of lines that performed superior to the parents. I tried my best to answer the question, it this truly pertaining to line 299.

301-376 It is unnecessary to use so many levels of confirmation for a simple QTL study (verified, confirmed, positionally confirmed + abbreviations). Important is only confirmation within the study (across environments) and confirmation across studies. QTL confirmed only within this study (i.e. Gm 9) get too little attention. Comment clearly understood and I refocus to highlight pertinent information gained from the study and simplified the language. You were correct that so many types of confirmation was confusing.
L315 I agree that confirmation across environments is the crucial criterion. But now the criterion became "both years". Does not make sense to me. Not intending to change the criterion but to also indicate the repeat identification of this QTL from a previous year, Language was clarified to indicate it was one year of three locations and one year reference to one location data set. Pointing out multiple year findings is important because it indicates a stable QTL, which are more effective for MAS.
L324 Please check citation style. Thank you, corrected.
L342 I don't agree. If you confirm a QTL in other donors you can deduce that the alleles of this QTL may not be too rare and that it functions in different genetic backgrounds. I do understand your comment and you make a valid point.

L383 "pleiotropy" Yes.

Thank you for your thoughtful and helpful comments on this review.

 

 

 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Still, in my opinion, the work is too poor to be published as an original article (rather as a short communication). This does not change the fact that it is a complete version. I leave the decision to the Main Editor.

Author Response

Reviewer 1

Still, in my opinion, the work is too poor to be published as an original article (rather as a short communication). This does not change the fact that it is a complete version. I leave the decision to the Main Editor.

We feel the work demonstrated in this research project does not represent poor research but instead valuable components of a puzzle that help us to assemble a more complete picture and pathway in pursuit of these important quality traits. Traits that the industry in deeply concerned about for efficiency and sustainability. Without QTL studies such as this, we cannot fully account for synthesis of the quality traits nor recognize the genetic components responsible. I hope this explanation helps to help your perspective on the potential value of our study.

Reviewer 2 Report

The authors improved the manuscript but at some instances it seems like they didn't take much time for the revision.
The following points still have to be addressed:

L78-86 This section still interrupts the results from previous studies. Please restructure.
L100-L103 If you distinguish verification from confirmation you have to define both terms beforehand. Isn't objective 2 the same as QTL verification (objective 1)?
L124 Still not apparent that the F5:7 rows were used for year one of the present study. Just add a few words to clear things up please.
L172 It's not important to me which style you use for formulas but it has to follow one of the many traditions. I suggest the following formula given that BTij is the interaction: Yij = μ + Bi + Tj + BTij + εij
L177 I can't see that you changed anything here. It still looks like you have used formula 1 to estimate adjusted means across environments for QTL analysis. Using this formula you will not get means for separate environments.
L204 RIL lines --> RILs
L239 Column Yield is empty. Please invest some time.
L242 Which study? This sentence was good in the original version.
L268 Typo
L323 Please make a more concise statement. It is not clear wether you talk about transgressive segregation here.
L325-401 Again: The differentiation between "confirmed QTL", "verified QTL" and "positionally confirmed QTL" is not a universal practice in the field of genetics. The fact that the definitions for these terms are scattered all over the discussion does not help either. More importantly, this differentiation is unnecessary for this study in my opinion. You should either (1) write a small paragraph that defines all three terms and makes clear why they are necessary for this study or (2) use the words confirmed and verified interchangeably (see my comment in review round 1).

Author Response

Reviewer 2

The authors improved the manuscript but at some instances it seems like they didn't take much time for the revision.
The following points still have to be addressed:

L78-86 This section still interrupts the results from previous studies. Please restructure. Thought was expanded and the paragraph was restructured to be more succinct.
L100-L103 If you distinguish verification from confirmation you have to define both terms beforehand. Isn't objective 2 the same as QTL verification (objective 1)? We will use confirmed QTL and identify the QTL that met specific Soybase criteria.

L124 Still not apparent that the F5:7 rows were used for year one of the present study. Just add a few words to clear things up please I have adjusted the language to bring more clarity.

L172 It's not important to me which style you use for formulas but it has to follow one of the many traditions. I suggest the following formula given that BTij is the interaction: Yij = μ + Bi + Tj + BTij + εij,  Thank you, the equation was determined to be unneccesary by co-authors and myself, it has been deleted for simplicity.  My apologies for the confusion.
L177 I can't see that you changed anything here. It still looks like you have used formula 1 to estimate adjusted means across environments for QTL analysis. Using this formula you will not get means for separate environments. The analysis was for all environments but also allowed us to find common QTL in all enviroments.
L204 RIL lines --> RILs. Corrected
L239 Column Yield is empty. Please invest some time. I missed that and thank you for pointing it out.  Seed weight was removed, also removing the correlation of seed weight to yield.
L242 Which study? This sentence was good in the original version.””In the current study” was added
L268 I think I corrected it
L323 Please make a more concise statement. It is not clear whether you talk about transgressive segregation here. I have clearly defined the lines as transgressive segregates and added language to most accurately describe them.
L325-401 Again: The differentiation between "confirmed QTL", "verified QTL" and "positionally confirmed QTL" is not a universal practice in the field of genetics. The fact that the definitions for these terms are scattered all over the discussion does not help either. More importantly, this differentiation is unnecessary for this study in my opinion. You should either (1) write a small paragraph that defines all three terms and makes clear why they are necessary for this study or (2) use the words confirmed and verified interchangeably (see my comment in review round 1). In discussions with the co-authors we took your comment seriously and agreed that was confusing.  We resolved to use confirmed QTL.