Genomic selection for agronomical phenotypes using genome-wide SNPs and SVs in pearl millet

Area of Science:

• Plant genetics and genomics
• Crop improvement and breeding
• Bioinformatics and statistical genetics

Background:

• Pearl millet is a vital, stress-resilient crop, but its breeding progress is hindered by limited use of molecular-assisted breeding.
• Existing breeding strategies often underutilize the full spectrum of genetic variation available in germplasm.

Purpose of the Study:

• To evaluate the utility of single nucleotide polymorphism (SNP) and structural variant (SV) markers in pearl millet genomic prediction.
• To compare the performance of SNP and SV markers in capturing genetic diversity and predicting phenotypes.
• To identify elite pearl millet accessions for accelerated breeding programs.

Main Methods:

• Collected 1,455,924 SNP and 124,532 SV markers from 242 pearl millet inbred accessions.
• Assessed 120 yield-related phenotypes.
• Evaluated nine genomic prediction models, comparing SNP and SV marker performance.

Main Results:

• SV markers revealed genetic diversity not captured by SNP markers.
• SV markers demonstrated higher heritability and prediction accuracy compared to SNP markers across various models.
• Genomic Best Linear Unbiased Prediction (GBLUP) with SV markers and Bayesian methods with SNP markers showed optimal performance, respectively.
• Identified eight candidate accessions using SNP and four using SV markers for high genomic estimated breeding values (GEBV).

Conclusions:

• SV markers provide a valuable complement to SNP markers for comprehensive genomic prediction in pearl millet.
• The study identified specific elite accessions, including 'P23' for panicle number, accelerating molecular breeding efforts.
• Findings support the integration of both SNP and SV marker data for enhanced pearl millet variety development.