GWAS on the Attack by Aspen Borer Saperda calcarata on Black Cottonwood Trees Reveals a Response Mechanism Involving Secondary Metabolism and Independence of Tree Architecture

Round 1

Reviewer 1 Report

Thank you for the interesting manuscript! 

Below ar my suggestions and comments.

I think Fig. 1 is not necessary.

You mention than some of the top SNPs are in intergenic regions in Table 2. Please elaborate on this in a sentence or two in the discussion. Do you believe transcription regulating motifs are affected?

In "Genotype effects" section the CC genotype is mentioned. It contradicts with Figure 5. Please fix it.

In Discussion, sentence "From our results, the relationship between genetic variation and the presence of wood borer attack seems weak since a reduced quantity of significant SNPs was detected." - please provide a comparison with another study where a strong relationship is shown for another factor. This sentence begs for additional explanations. To avoid going into details about why certain thresholds have been selected, I think it would be sufficient if you would provide reference to another study where a strong relationship was detected due to detecting lots of significant SNPs.

Improve column widths in Table 2.

As you explain, Chr10_22235609 is located in the intronic region. But in Table 2 you indicate an amino acid change. This change would not happen as it is located in an intron. Please fix this.

And, maybe most importantly, I didn't notice information about where the raw data (SNP table and SRA data) can be found.

Please explain the meaning of MAF in section "Genotyping and SNP data processing".

Consider writing the sentence "The analysis of the phenotypic association between the presence of insect attack and the number of stems indicated that the distribution of the attack was independent of the distribution of the number of stems (p-value = 0.244, Chi2 test)." in the form "The analysis of the phenotypic association between the presence of insect attack and the number of stems indicated that the distribution of the attack was independent of the number of stems (p-value = 0.244, Chi2 test). ".

Consider deleting sentence "Figures 3A and 4A show the LD heatmaps for the significant markers about markers in genes located in the surrounding regions.".

In LD analysis section, when mentioning gene Potri.003G110200.1, indicate that it is not depicted in the figure.

In discussion - to induce denfensives responses against pathogenic - I suggest using "defensive".

Table 1. abbreviate meters as m and use kg, not Kg.

In Discussion, consider substituting "near genes" with "genes in proximity".

Author Response

Please see the attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

The aspen borer (Saperda calcarata) poses a significant threat to cottonwoods. In this study, the authors analyzed the genetic mechanisms underlying the response of black cottonwood to S. calcarata infection in a common garden trial and assessed its influence on wood architecture traits and metabolite production. This paper is well written and a rather easy read which is always good. That said, I have several issues with it. Below, I have listed some of the main comments:

 1. Introduction

The introduction of the article is logically clear, although in some cases, the level of detail provided on certain results may not be necessary. However, for important points that are of significant relevance to the article, a single citation may not be sufficient to fully convince the reader. Such as: Hence, with the destruction of just a single terminal shoot, the growth form of the tree can be permanently and dramatically altered from single-stemmed to multistemmed (Chamberlin and Aarseen, 1996).

2. Materials and Methods

-

I think this is your previous research result, and here the author should provide details on the number of genotypes used for the metabolomics analysis, the sampling details, and whether or not replicates were performed.

-

The phenotype assessment during the second growing season was conducted in which month, during the period of infestation or after the infestation, or after the cessation of tree growth? This information is important for accurate phenotype measurement.

-

Using only the presence or absence of attack as a phenotype makes the article overly simplistic. Since metabolic analysis was conducted, why was there no GWAS analysis performed on significant metabolites such as organic compounds and phenolic substances? Is there a correlation between their differences and the presence or absence of attack?

-

The manuscript mentioned that "The modeling included the number of stems and the relative abundance of individual metabolites as covariates", and "The corresponded to 101 provenances representing part of the P. trichocarpa natural distribution range." When conducting GWAS analysis, the influence of population structure should be considered and potential population structure errors should be controlled through principal component analysis or population structure adjustment.

-

The manuscript does not present any enrichment analysis charts on Gene Ontology; rather, it seems to only provide gene annotation results.

3. Results

-

How could there be no statistical results of the phenotype data in the GWAS analysis? At least, the statistical results of each genotype and whether they conform to the normal distribution should be presented in the supplementary materials.

-

The determination of the threshold for GWAS analysis should be added to the corresponding section of the Materials and Methods. Why were only the top 10 SNPs retained instead of the ones that reached the significant threshold? How were the genes annotated, using software such as snpEff or by looking for upstream and downstream regions based on LD? These details should be included in the Materials and Methods section. The abstract and introduction should also mention how many SNPs were identified through GWAS and how many gene models were associated.

-

Performing haplotype analysis in combination with phenotype data is meaningful, and this should be presented together with the results of individual SNP analysis in the LD analysis section, instead of being presented separately.

4. Discussion

-

Like the introduction, the discussion is also commendable. Some of the content is too redundant and needs to be streamlined. For example, the third paragraph, etc.

-

Final, based on the statistical results of this study, if some genotypes with higher resistance to Saperda calcarata can be provided, it would be more informative for the demonstration and promotion of research on resistance to the longhorn beetle.

Author Response

Please see the attached file.

Author Response File: Author Response.pdf