GS4PB: An R Shiny application to facilitate a genomic selection pipeline for plant breeding

Area of Science:

• Agricultural Science
• Genetics
• Bioinformatics

Background:

• Genomics-assisted breeding is crucial for increasing crop yields amid climate change and resource limitations.
• Small and medium-sized breeding programs face challenges integrating genomic selection due to a lack of comprehensive software tools.
• Bridging this gap is essential for accelerating genetic gain in plant breeding.

Purpose of the Study:

• To develop and present an integrated, open-source software tool for streamlining the genomic prediction pipeline in plant breeding.
• To lower the barrier for entry into advanced genomic prediction techniques for breeders.
• To demonstrate the efficacy of the developed tool in enhancing genetic gain.

Main Methods:

• Developed an R Shiny web application, GS4PB (Genomic Selection For Plant Breeding), integrating open-source tools.
• The pipeline includes genotypic data processing (filtering, imputation), data merging, enviromics covariate collection, kinship estimation, training set optimization, and cross-validation.
• Implemented genomic prediction for single/multiple traits across environments.

Main Results:

• GS4PB successfully integrates essential steps of the genomic prediction pipeline.
• An experiment comparing phenotypic and genomic selection demonstrated that genomic selection performed comparably to phenotypic selection for advancing breeding lines.
• The tool facilitates the application of advanced genomic prediction methods.

Conclusions:

• The GS4PB application provides a user-friendly, integrated solution for genomic prediction in plant breeding programs.
• This tool democratizes access to advanced genomic selection technologies, empowering breeders to accelerate genetic gain.
• Widespread adoption of GS4PB can significantly contribute to meeting global food production challenges.