PubMed · 2026-02-18
Scientists are building comprehensive genetic maps of rice that capture far more diversity than a single reference genome, revealing hidden genes and variations that control yield, disease resistance, and stress tolerance. The challenge now is turning this research goldmine into practical tools that plant breeders can actually use.
Rice pangenomes have uncovered extensive structural variations, presence/absence variations (PAVs), and entirely novel genes not found in any single reference genome, expanding the known genetic diversity of the crop.
AI and machine learning show strong potential for interpreting complex pangenomic data and accelerating trait discovery through genomic selection, but are currently limited by high computational demands and lack of breeder-friendly interfaces.
Key barriers preventing pangenomic data from entering routine breeding pipelines include complex graph-based data structures, difficulty detecting multiallelic variants from population-wide sequencing, and the absence of practical genotyping tools for breeders.
