Diversity Characterization of Soybean Germplasm Seeds Using Image Analysis
, Jeong Won Jo 3, Xiaohan Wang 1, Myoung-Jae Shin 1, On Sook Hur 1, Bo-Keun Ha 2,* and Bum-Soo Hahn 1,*Round 1
Reviewer 1 Report
The experiments are well designed, although modern approaches in machine vision and pattern recognition are notably absent; I believe automatic feature extraction from the segmented seeds could potentially improve the results.
I also suggest the authors to review their references since only one fourth of them are from recent years.
Sections 3.2 and 3.3 are presented as literal descriptions of the results in the tables, thus providing little information. I suggest the authors to summarize the information contained in there.
There are several values that are presented without their corresponding units, most notably in table 2.
Author Response
Response to Reviewer 1 Comments
Point 1: I also suggest the authors to review their references since only one fourth of them are from recent years.
Response 1: I strongly agree with the reviewer's comments. Unfortunately, not many good publications have been made in the recent years in the field of seed phenome research. There are publications made in phenome research in the recent years, but they would not be completely relevant to the present data presented.
Point 2: Sections 3.2 and 3.3 are presented as literal descriptions of the results in the tables, thus providing little information. I suggest the authors to summarize the information contained in there.
Response 2: Thank you very much for your suggestion. We have rewritten parts of the 3.2 and 3.3 that were descriptions of the results presented in the tables/figures.
Point 3: There are several values that are presented without their corresponding units, most notably in table 2.
Response 3: Thank you to the reviewer for your comments. The missing units have been added to Table 2. 100 Weight (g), Height (mm), width (mm), Perimeter (mm) and Area (mm) were inserted in units respectively. On the other hand, AR, Solidity, Circularity, and Roundness are still omitted because they do not have units.
Reviewer 2 Report
General comments
The paper entitled “Diversity Characterization of Soybean Germplasm Seeds Using Image Analysis” treats about a topic of the highest interest in the agricultural scope. The main aim of the Authors consists in characterizing the Soybean Germplasm by using image analysis. The abstract summarizes in short the content and outlines well the aims of the contribution. Keywords are appropriate. The highlights are missing; therefore this reviewer cannot judge them.
Introduction outlines the approach put in place by Authors to reach out their aims. In particular, Authors stress the role of Orthogonal projections to latent structures discriminant analysis (OPLS-DA) to describe effectively the results of the characterization performed by means of the cluster analysis. That methodology is recent but not novel and can have some drawback as highlighted by Worley, B., & Powers, R. (2016). PCA as a practical indicator of OPLS-DA model reliability. Current Metabolomics, 4(2), 97-103. In fact, that method aggressively forces separations between experimental groups. The first question to Authors concerns if they have put the needed attention for not fall in that pitfall. Another curiosity, have Authors checked the Bartlett test and the needed assumptions before applying PCA? Anyway, the manuscript is interesting and well structured. The experimental design is sufficiently good.
References are abundant and recent. Figures and graphics are clear and of good quality. The manuscript content then falls without doubt under the journal scope. Used language is overall fluent and grammatically correct.
Author Response
Point 1: The highlights are missing; therefore this reviewer cannot judge them.
Response 1: Thank you for your comment. We are sorry, we did not include highlights section in the current manuscript as it was not mandatory for submission.
Point 2: Introduction outlines the approach put in place by Authors to reach out their aims. In particular, Authors stress the role of Orthogonal projections to latent structures discriminant analysis (OPLS-DA) to describe effectively the results of the characterization performed by means of the cluster analysis. That methodology is recent but not novel and can have some drawback as highlighted by Worley, B., & Powers, R. (2016). PCA as a practical indicator of OPLS-DA model reliability. Current Metabolomics, 4(2), 97-103. In fact, that method aggressively forces separations between experimental groups. The first question to Authors concerns if they have put the needed attention for not fall in that pitfall. Another curiosity, have Authors checked the Bartlett test and the needed assumptions before applying PCA? Anyway, the manuscript is interesting and well structured. The experimental design is sufficiently good.
Response 2: The Bartlett sphericity test and the Kaiser-Meyer-Olkin (KMO) test were checked when using XLSTAT for principal component analysis. The results show that the P value of Bartlett's test rejected null hypothesis, that is, it is considered that the research data can be extracted by principal components, and the hypothesis 2 is satisfied. In addition, the KMO test coefficient is greater than 0.6, indicating that the sample meets the requirements of a reasonable data structure. Corresponding text has been added to the manuscript.
