Dissecting the genetic basis of agronomic traits by multi-trait GWAS and genetic networks in maize (Zea mays L.).

Agronomic traits in maize (Zea mays L.) are complex and modulated by pleiotropic loci and interconnected genetic networks. However, the traditional single-trait genome-wide association study (GWAS) method often misses genetic associations among traits, overlooks pleiotropic effects, and underestimates shared regulatory mechanisms. In the current study, we employed multi-trait analysis of GWAS (MTAG) and constructed a genetic network to dissect the genetic architecture of 18 agronomic traits across a genetically diverse panel of 2,448 maize inbred lines. Incorporating MTAG significantly improved the detection of pleiotropic loci that had not been detected by single-trait GWAS. Using a genetic network, we uncovered numerous previously unrecognized connections among traits related to plant architecture, yield, and flowering time. The 49 detected hub nodes, including Zm00001d028840 and Zm00001d033859 (knotted1), influence multiple traits. Co-expression analysis of candidate genes across two developmental stages validated their distinct yet complementary roles, with Zm00001d028840 linked to early cell wall remodeling and Zm00001d033849 involved in chromatin remodeling during tasseling. Moreover, we integrated results from GWAS, MTAG, and genetic network analyses to prioritize pleiotropic loci and hub genes that regulate multiple agronomic traits. This integrative approach offers a practical framework for selecting stable, multi-trait-associated targets, thereby supporting more precise and efficient crop improvement strategies. Supplementary Information: The online version contains supplementary material available at 10.1007/s42994-025-00241-4.