A Forward Genetics Strategy for High-Throughput Gene Identification via Precise Image-Based Phenotyping of an Indexed EMS Mutant Library.

Abstract

Ethyl methanesulfonate (EMS) mutants are widely used for genetic analysis; however, EMS-derived mutant populations are not amenable to traditional genome-wide association studies (GWAS) because the EMS mutations are present at extremely low frequencies. To address this challenge, this work develops the GeneHunter-Gene-Level Association (GH-GLA) pipeline using an EMS-generated population of wheat (Triticum aestivum) mutants and an image-based phenotyping platform. GH-GLA enables comprehensive exploration of phenotypic variation induced by genome-wide saturation mutagenesis. Using GH-GLA to quantify 83 traits in the wheat population reveals that variation in spikelet geometry is significantly associated with key agronomic traits, including thousand-kernel weight. Using this indexed wheat EMS population and phenotype data, GH-GLA identified 5905 genes that are significantly associated with specific traits. Analysis of knockouts generated by gene editing, together with haplotypes affected by selection during breeding and genetic variation in 262 wheat accessions, confirm the roles of TaAN-1, TaBAM5L, and TaXTH28L in regulating thousand-kernel weight and spikelet angle. Furthermore, this work establishes an epistatic interaction network between gene pairs to elucidate their combined effects on the phenotype. Overall, GH-GLA provides a powerful strategy for functional gene identification, and the alleles discovered here offer valuable genetic resources for crop improvement.

Code and data availability

The paper's GH-GLA analysis code is publicly available on GitHub with explicit availability language. The phenotypic data (OMIX010498) and VCF data (GVM000963) are deposited in repositories whose URLs are not in the allowed list, so they cannot be cited as assets here.