Haplotype associated RNA expression (HARE) improves prediction of complex traits in maize

Area of Science:

• Genomics
• Quantitative Genetics
• Bioinformatics

Background:

• Genomic prediction commonly uses single-site polymorphisms, limiting insights into haplotype function and biological mechanisms.
• Current methods struggle to capture the combined effects of alleles within haplotypes, impacting prediction accuracy for complex traits.

Purpose of the Study:

• To introduce a novel, cost-effective method for imputing cis haplotype associated RNA expression (HARE).
• To assess the transferability of HARE estimates across different tissues and populations.
• To evaluate the performance of genomic prediction models using HARE estimates for complex traits in maize.

Main Methods:

• Developed and applied a method to impute cis haplotype associated RNA expression (HARE), focusing on cis-acting variants near genes.
• Assessed HARE transferability across diverse maize panels (Goodman Association panel, Nested Association Mapping panel) and tissues.
• Incorporated HARE estimates into genomic prediction models for 26 complex traits, comparing accuracy against control methods.

Main Results:

• HARE estimates demonstrated higher transferability across tissues and populations compared to measured transcript expression.
• HARE captured approximately one-third of the variation in gene expression.
• Genomic prediction models using HARE achieved up to 15% higher accuracy than control approaches, particularly when trained and tested across diverse panels.

Conclusions:

• Imputing RNA expression by haplotype (HARE) is a stable, cost-effective, and transferable approach for genomic prediction.
• HARE provides functional information beyond haplotype structure, enhancing prediction accuracy for complex traits.
• The HARE method offers a promising avenue for improving genomic selection strategies in plant breeding.