Tasselyzer, a machine learning method to quantify maize anther exertion, based on PlantCV.

Area of Science:

• Agricultural Science
• Plant Biology
• Genetics

Background:

• Maize anther exertion is crucial for male fertility and is influenced by genetics and environment.
• Manual quantification of anther exertion is subjective and labor-intensive, hindering research and breeding efforts.

Purpose of the Study:

• To develop and validate Tasselyzer, an accessible, cost-effective, and time-saving image-based program for quantifying maize anther exertion.
• To provide a quantitative and automated alternative to traditional manual scoring methods.

Main Methods:

• Utilized the PlantCV platform for image analysis of maize tassels.
• Developed an algorithm to quantify anther exertion based on regional color differences within the tassel.
• Applied Tasselyzer to 22 maize lines across six genotypes.

Main Results:

• Tasselyzer demonstrated high precision with an F1 score greater than 0.8.
• The method proved effective in capturing regional differences in anther exertion within the tassel.
• Parameter customization for specific maize lines was shown to be straightforward and practical for improving precision.

Conclusions:

• Tasselyzer offers a valuable, automated, and efficient solution for assessing maize anther exertion.
• This tool supports maize research and breeding programs by enabling rapid and objective male fertility evaluations.
• The image-based approach enhances the accuracy and efficiency of quantifying a key trait in maize.