Conventional and hyperspectral time-series imaging of maize lines widely used in field trials

ABSTRACT

**Reviewer 2. Cory Hirsh **

This manuscript describes the generation of a time-series dataset of conventional and hyperspectral images of commonly known and important maize lines. The authors describe the methods of data collection and how it is useful, especially in conjunction with other already available datasets for the same lines. The authors begin to analyze the dataset generated, focusing on biomass measures and determining heritability. The authors conclude that they believe it is important and necessary to combine controlled environment data with field data to tackle problems facing crop production. I do have several comments about the manuscript in its current form:

My main concern about the manuscript is the amount of data use in the article. The manuscript was submitted as a 'Data Note', but it is not obvious this data is exceptional, rare, or novel as it was collected nearly 2 years ago. One criteria to review this type of article is dataset size. The authors are claiming a dataset size of ~500Gb, but this includes data (thermal infrared and fluorescence images) that was not mentioned in the manuscript except that it was collected. I applaud the authors for the willingness to be so open with their data, but I'm not convinced that one month worth of images for 32 genotypes is enough for publication.
The manuscripts main point is not to get into conclusions based on their image analysis, but I would have liked to have seen more strenuous ground truthing. The manual measurements were made only at the very last time point. These really should encompass the variation of plants throughout development. How can we determine if the measured traits are accurate at day 9 for example? Nothing can be done for true manual measurements, but digital manual measurements could be made and correlated with image analysis extracted values.
Board sense heritability needs to be corrected throughout the manuscript.

Re-review:

This manuscript describes the generation of a time-series dataset of conventional and hyperspectral images of commonly known and important maize lines. The authors describe the methods of data collection and how it is useful, especially in conjunction with other already available datasets for the same lines. The authors begin to analyze the dataset generated, focusing on biomass measures and determining heritability. The authors conclude that they believe it is important and necessary to combine controlled environment data with field data to tackle problems facing crop production.

Comments: I want to clarify my first review of this manuscript. It was not my intention to make it seem as the dataset generated for this manuscript is not important, large, or useful for the broader maize and plant phenotyping community. This dataset could be very useful for some research groups, including the corresponding authors group. The authors response to the age question of the dataset of, look at the cycle time of data collection to publication in plant phenomics is generally longer, I totally agree with. The authors give numerous examples to back up this point. I'm not disputing this, but the authors should also note the amount of downstream analyses and new biological findings that are in these manuscripts as well. The importance of the presented dataset as outlined by the authors is its ability to link with other already available datasets, which isn't shown in the manuscript. This paper is a data release paper with a valuable, controlled, and well documented dataset. The real value in the dataset will be shown in subsequent publications that begin to combine the multiple datasets available from these maize lines (field phenotyping, genotyping, controlled environment phenotyping).

Read the original source

Now published in GigaScience doi: 10.1093/gigascience/gix117

Zhikai Liang 1Center for Plant Science Innovation, Department of Agronomy and Horticulture, University of Nebraska-Lincoln, Lincoln, 68503, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteORCID record for Zhikai LiangPiyush Pandey 2Department of Biological System Engineering, University of Nebraska-Lincoln, Lincoln, 68503, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteVincent Stoerger 3Plant Phenotyping Facilities Manager, University of Nebraska-Lincoln, Lincoln, 68503, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteYuhang Xu 4Department of Statistics, University of Nebraska-Lincoln, Lincoln, 68503, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteYumou Qiu 4Department of Statistics, University of Nebraska-Lincoln, Lincoln, 68503, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteYufeng Ge 2Department of Biological System Engineering, University of Nebraska-Lincoln, Lincoln, 68503, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteJames C. Schnable 1Center for Plant Science Innovation, Department of Agronomy and Horticulture, University of Nebraska-Lincoln, Lincoln, 68503, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteORCID record for James C. Schnable

A version of this preprint has been published in the Open Access journal GigaScience (see paper https://doi.org/10.1093/gigascience/gix117 ), where the paper and peer reviews are published openly under a CC-BY 4.0 license.

These peer reviews were as follows:

Reviewer 1: http://dx.doi.org/10.5524/REVIEW.100930

Read the original source

Conventional and hyperspectral time-series imaging of maize lines widely used in field trials

This article has been Reviewed by the following groups

Listed in

Abstract

Article activity feed