The digital map of homologous gene clusters across Poaceae species unveils the complex evolutionary trajectories of genes in Triticeae.

Gene origin, duplication, and loss are key drivers that shape genome evolution, phenotypic diversification, and plant adaptation. Nevertheless, the fine-scale evolutionary trajectories of genes within their local genomic contexts across diverse genera remain poorly characterized. Here, we developed an approach, GoldMiner, that uses homologous gene clusters (HOCs) as evolutionary units for investigating gene evolution and enables hierarchical alignment of pan-genomic HOCs across species and genera. We constructed a genus-level pangenomic map of half a million indexed HOCs across 248 diploid genomes from 25 Poaceae species. We found that most newborn HOCs were derived from existing homologous HOCs. The newborn HOCs in Triticeae were strongly associated with stress response and defense pathways. Two-thirds of Triticeae HOCs, including those containing NLR genes, underwent substantial expansions during evolution, contributing to disease resistance in wheat and barley. Genomic redundancy between wheat subgenomes drives HOC loss associated with genetic variation of wheat populations. We dissected seed storage protein (SSP) gene evolution at two scales: inter-HOC turnover, which governs the origin and amplification of distinct loci across Triticeae, and intra-HOC divergence, which drives functional diversification among paralogous copies. Finally, an interactive web platform, waGOLD (https://wheat.cau.edu.cn/TGT/waGOLD), was developed for the community to explore evolutionary trajectories of HOCs. Overall, we presented a digital atlas of gene evolution for Poaceae species as a resource that opens new avenues for fine-scale gene family evolution and provides a practical framework for constructing genus-level gene-based pangenomes.