ShinyFruit: interactive fruit phenotyping software and its application in blackberry.

Area of Science:

• Horticultural Science
• Plant Breeding
• Bioinformatics

Background:

• Horticultural plant breeding requires extensive phenotypic data for quality assessment.
• High-throughput phenotyping pipelines are crucial for advanced breeding strategies like genome-wide association studies and genomic selection.

Purpose of the Study:

• To introduce ShinyFruit, an R-based web application for image-based phenotyping of fruit and vegetable traits.
• To evaluate ShinyFruit's accuracy in measuring blackberry size and red drupelet reversion (RDR) compared to traditional methods.

Main Methods:

• Developed an R-based web application, ShinyFruit, for automated image analysis.
• Compared ShinyFruit's measurements of blackberry length, width, and RDR with ImageJ and manual measurements.
• Utilized a population of blackberry cultivars and breeding selections.

Main Results:

• ShinyFruit showed strong positive correlations with manual measurements for blackberry length (r = 0.96).
• High correlation was observed between ShinyFruit and ImageJ for RDR estimates (r = 0.96).
• Weaker correlations were found for manual RDR estimation (r = 0.62–0.70); no genotypic differences in width were detected.

Conclusions:

• ShinyFruit provides a viable, open-source solution for efficient phenotyping of size and color in blackberry fruit.
• The application's adjustable settings enhance its utility for diverse fruits and vegetables.
• Further studies may strengthen correlations by maximizing genotypic variance for traits like RDR.