Designed diversity: from marker-assisted backcrossing to computationally optimised polygenic introgression

Crop breeding strategies for trait introgression are shifting from empirical backcrossing, focused on a small number of target loci, to a predictive, large-scale, design-driven approach. This approach frames the introgression of desirable trait loci as a constrained multi-objective optimisation problem, balancing success probability, elite genome recovery, time, and resource use. Computational tools such as predicted and expected cross value, and look-ahead selection, prioritise crosses and support multi-generation planning. Advances in genomic resources and breeding accelerators, including pan-genomes, introgressiomics, speed breeding, and gene editing, expand the design space for precise allele stacking. By embedding these technological innovations within predictive breeding frameworks, breeders can translate genomic insights into practical strategies to refine introgression pipelines, transforming trial-and-error programmes into 'designed diversity' pipelines for resilient, high-yielding crops.