Challenges in Bringing Pangenome Research into Breeding: A Case Study in Rice

Crop breeding has entered the pangenomics era, unlocking a far more comprehensive view of genetic diversity than a single reference genome can capture. In rice (Oryza sativa), a staple crop critical to global food security, the construction of pangenome resources has uncovered extensive structural variations (SVs), novel genes, and presence/absence variations (PAVs) that underpin key agronomic traits. As the rice pangenome matures from a research resource into a practical breeding tool, it promises to accelerate the development of higher-yielding, stress-resilient, and disease-resistant varieties. This transition represents a pivotal advance toward sustainable agriculture and enhanced global food security, while also establishing a model for applying pangenomics to other crops. Here, we review how rice pangenome research, encompassing both cultivated and wild species, has advanced trait discovery (from yield improvement and disease resistance to stress tolerance) and enabled new molecular breeding strategies. Despite these advances, several challenges remain before pangenomic data can be routinely integrated into breeding pipelines. The complexity of graph-based data structures, difficulties in detecting multiallelic variants from population-wide resequencing data, and the lack of breeder-friendly genotyping tools are significant barriers. Additionally, while artificial intelligence (AI) and machine learning (ML) approaches show great promise for interpreting complex pangenomic data and accelerating trait discovery, their practical adoption is hindered by the absence of breeder-oriented interfaces, integration challenges with phenotypic and environmental data, and high computational demands. Overcoming these issues will require interdisciplinary collaboration, robust infrastructure, and innovations focused on practical breeding needs across diverse crop species.