Mylnefield Trust Scholars: Revolutionizing Raspberry Breeding: Advancing into the Molecular Era with Cutting-Edge Molecular Marker Tools

Climate change, arising pests and diseases, and labour costs are straining many local agricultural systems. Advances in genetics and genomics research have successfully identified molecular markers associated with increased yield, resistance to diseases or timing of key phenology events. A key challenge for crop breeding is the time and resources required for marker-assisted selection.

This exciting multidisciplinary project will combine machine learning, quantitative and population genetics to advance genomic selection (GS) in raspberry breeding (Rubus idaeus L). This project provides exceptional training value in state-of-the-art genomics with direct deployment into real-world breeding scenarios for industry with immense potential to learn the skills needed for future scientific breakthroughs and advancing molecular breeding.

The breeding method, GS, uses genomic prediction models to select individuals based on estimated phenotype and reduces significantly the time and cost required to develop a cultivar, allowing a fast transference of alleles of interest into commercial breeding pipelines.

Aim: Implement GS in commercial raspberry breeding pipelines.

1)    Genome assembly and GWAS (genome-wide association study)

The soft fruit genetics group of the James Hutton Institute has developed an association mapping panel with input from Biomathematics and Statistics Scotland. The panel is an extensively characterised collection of diverse raspberry genotypes designed for GWAS that will allow the student to:-

a)    use detailed datasets to curate phenotypic data and perform analysis of the heritability of the traits to select the most suitable for GS pipelines.

b)    carry out population analysis.

c)     define and characterise a training set, a subset of lines used to produce the genomic model.

d)     training in trait analyses used in commercial raspberry breeding programmes with soft fruit breeders at James Hutton Ltd (JHL).

2)    Development and validation of GS models

The student will learn:-

a)    how mapping genotypic data from the panel to a new genome assembly is performed with involvement in the assembly and annotation process.

b)    Perform GWAS over the panel to identify loci associated with traits of interest.

c)    identify associated molecular markers and test in commercial material with the Molecular Diagnostics Unit of JHL and gain valuable experience of using commercial molecular markers.

d)    how to build statistical models to produce genomic predictions of traits of interest, using e.g. GWAS-informed parametric models and non-parametric machine learning based methods.

e)    Evaluate the performance of the models using widely used cross validation methods e.g. leave-one-out.