Near Infrared Spectroscopic Evaluation of Starch Properties of Diverse Sorghum Populations

Round 1

Reviewer 1 Report

This article is useful for readers. There are some problems that have to be addressed before it can be published.

1. In the introduction, the statements on sorghum and NIR are too few. Actually, there are many publications that have reported. So, the authors should address it. And give readers the reasons why you did this study.
2. Increase the relevant information of the near-infrared spectrometer used in this work and the measurement parameters for samples;
3. What are the differences between AP1, AP2, AP3, and AP4?
4. Please add a figure containing the original spectra,
5. Eliminate the section “3.6. Relationship between, starch, amylose and protein contents in grain sorghum populations” because it doesn’t have something to do with the theme of this article.
6. Part 3.7 is too verbose.

Author Response

1. In the introduction, the statements on sorghum and NIR are too few. Actually, there are many publications that have reported. So, the authors should address it. And give readers the reasons why you did this study.

We have indicated that NIR is used for measurement of numerous quality traits of cereal grains including sorghum. While the reported NIR methods for sorghum grain analysis mostly use ground samples, the literature on NIR methods for intact grain sorghum is very limited. For evaluation of small samples from breeding trials, importance of the use of intact grains was highlighted and therefore we have shown the need for and importance of developing NIR method for starch/Amylose analysis in intact sorghum grains. We have also cited two excellent reviews on the use of NIR for screening cereal breeding materials and use of NIR for evaluation of starch in cereals and starchy food.

2. Increase the relevant information of the near-infrared spectrometer used in this work and the measurement parameters for samples;

We added the fact that the instrument is a fixed resolution diode array spectrometer to distinguish it from FT spectrometer. As regards the measurement parameters, we recorded the spectra of samples only and the collected spectra and the reference laboratory starch and amylose data of respective samples were used the develop the starch and amylose calibrations.

3. What are the differences between AP1, AP2, AP3, and AP4?

These are the names given for the four different populations we used for amylose calibration developed. These were named in the same manner as used for starch sample populations as given in the text (lines 107-110).

4. Please add a figure containing the original spectra,

A figure with typical spectra from each sorghum population is added.

5. Eliminate the section “3.6. Relationship between, starch, amylose and protein contents in grain sorghum populations because it doesn’t have something to do with the theme of this article.

It is true that measurement of the protein is irreverent when NIR method is developed for starch analysis. However, protein is the second most important nutrient after starch in grain sorghum and protein content is also important in assessing varieties for specific purposes. For example, a variety with certain starch and protein levels may be more suitable as a feed ingredient than a variety with a higher starch content or a variety with relatively low starch content but with high protein and amylose levels may yield high ethanol yield and high protein DDDG than a high starch variety, making it more profitable for ethanol industry to improve profits when the ethanol yield and the quality of byproduct DDGS is considered. This kind of analysis of varieties for profit maximization is possible when both starch composition and protein levels are measured for varieties. We already have developed calibration for protein, so measurement of proteins provides more information to evaluate varieties. We were also able to highlight the relationship between protein and starch levels in sorghum populations. Therefore, we believe this section is valuable for the paper.

6. Part 3.7 is too verbose.

In these three short paragraphs we discuss how the NIR method can be used for screening breeding materials and show how spectroscopic data have been used by other researchers for evaluation of genetic potential of breeding materials.

Reviewer 2 Report

I suggest the to accept the paper.

Author Response

There is no comment from reviewer 2.

Reviewer 3 Report

Overall a well prepared manuscript presenting  interesting results. The introduction would benefit from small extension in some parts it is quite shallow. Methods are described accurately, but the description of material should be more clear. Discussion is lead well. Conclusions should be more realistic in respect to obtained results (same applies to abstract).

Line 42-49 using megatonne instead of Mg would be more profitable for reads as it is more commonly used

Line 55 amylose (singular)

Line 56 amylopectin (singular)

Line 54-62 whole paragraph is quite shallow, specify why waxy varieties are needed, where more starch is preferred, etc.

Line 64-66 as above, quite shallow and not taking into concidaration that methods for determination of amylose/amylopectin ratio are usually not accurate and providing different results

Line 87-118 I suggest adding a summary chart of the samples

Line 146 colorimetrically rather than calorimetrically

Line 146-167 why such an inaccurate method was used? Determination with iodine is rather for apparent measurements. Megazyme has alos kit amylose/ratio that can be used in four samples as well.

Line 426-431 Conclusions need to be more realistic as the R² for both starch and amylose for laboratory determined and predicted contents were below 0.9.

Author Response

1. Line 42-49 using megatonne instead of Mg would be moreprofitable for reads as it is more commonly used

Megaton is usually used for measurement of explosive power in terms of million tons of TNT explosive. However, for grains metric ton is used to express weight of grains. Since there are different expressions of tons (Imperial and US/long or short tons etc.) use of the standard SI unit megagram (Mg) which is equivalent to 1000 kilograms is the most suitable. To avoid any confusion, we used megagram at the first instance to define the acronym.

2. Line 55 amylose (singular)

Corrected

3. Line 56 amylopectin (singular)

Corrected

4. Line 54-62 whole paragraph is quite shallow, specify why waxy varieties are needed, where more starch is preferred, etc.

Just the starch content alone is not a good indicator of the suitability of a variety for a specific purpose as the amylose levels affect the utilization of the starch for the specific purpose. Therefore, this paragraph was revised giving an example as to why both starch content and amylose levels are needed to evaluate varieties for specific uses.

5. Line 64-66 as above, quite shallow and not taking into consideration that methods for determination of amylose/amylopectin ratio are usually not accurate and providing different results

We noticed that the starch chemistry part is not deep. Considering this paper is more on the NIR calibration side, we think brief introduction is OK.

Regarding the methods for amylose analysis, we reviewed all the possible methods, consulted a few starch chemistry experts in the field, and evaluated a few methods in our lab. Although the colorimetric method derived from amylose’s iodine binding capacity is not the best one for real amylose theoretically, it is more related with its functionality in real food systems. The procedure we adopted used trichloroacetic acid to stabilize the blue color and avoid the effect of protein on the absorbance. We obtained more consistent amylose results from this procedure on reference starch samples than from others including the Megazyme kit.

6. Line 87-118 I suggest adding a summary chart of the samples

A summary of the sample populations is added as Table 1.

7. Line 146 colorimetrically rather than calorimetrically

Corrected

8. Line 146-167 why such an inaccurate method was used? Determination with iodine is rather for apparent measurements. Megazyme has also kit amylose/ratio that can be used in four samples as well.

As stated above (response to question in lines 64-66), we evaluated several methods including the Megazyme kit for amylose/amylopectin. This is the most practical (for handling large number of samples) and more consistent method. The Megazyme kit is not only very time consuming, but also not applicable for samples with very low amylose content (its COA states applicable for amylose range in 5-95%). Because of this, we settled on the iodine binding method as a balance between sample throughput and representing functionality in foods. The method we used can be adapted to measure total amylose by adding an ethanol precipitation step.

9. Line 426-431 Conclusions need to be more realistic as the R for both starch and amylose for laboratory determined and predicted contents were below 0.9.

R² value is important to evaluate the suitability of a NIR model but the most important statistic is the RMSEC for calibrations and RMSEP for validations which shows how accurately or closely the interested trait can be predicted. R² can be improved by adding more samples and increasing the constituent range but not the prediction errors. Considering the prediction errors of these starch and amylose calibrations, these can be used for prescreening early generation materials in breeding programs as it impractical to use standard methods for evaluation of all early generation materials. We missed to indicate this fact in the conclusions and revised it indicating that these calibrations can be used to prescreen materials.

Round 2

Reviewer 3 Report

Although not all raised concerns were addressed as intended. The major aspects were corrected in a manner allowing for recommendation of the manuscript for publication.

Author Response

Please see the attachment.

Author Response File: Author Response.docx