Single-cross prediction with imputed multiomic data: a case study in rapeseed.

Area of Science:

• Plant genomics
• Agricultural science
• Bioinformatics

Background:

• Genomic prediction utilizes high-resolution plant genome data and phenotypic information for genotype selection in breeding programs.
• Single nucleotide polymorphism (SNP) arrays capture limited genomewide diversity, prompting exploration of whole-genome sequencing (WGS) and gene expression data imputation.

Purpose of the Study:

• To evaluate the impact of imputed whole-genome sequencing markers and expression data on genomic prediction accuracy in rapeseed hybrid breeding.
• To compare the effectiveness of SNP arrays versus combined genomic and expression data for predicting hybrid performance.

Main Methods:

• Utilized SNP arrays, whole-genome sequencing (WGS), and RNA sequencing on a rapeseed hybrid breeding population.
• Imputed marker and gene expression data across the population.
• Applied genomic prediction to estimate general and specific combining ability effects for untested hybrids.

Main Results:

• Imputed WGS markers and expression data increased marker density and linkage disequilibrium.
• Prediction accuracy for general and specific combining ability did not improve with the addition of imputed WGS and expression data compared to SNP array data alone.
• Potential reasons for lack of improvement include data redundancy, imputation errors, or environmental effects on gene expression.

Conclusions:

• SNP arrays are currently sufficient and reliable for genomic prediction in rapeseed hybrid breeding programs.
• Further research may be needed to overcome challenges associated with imputation accuracy and environmental influences on gene expression for enhanced genomic prediction.