Precision Plant Breeding: Integrating Genome Editing with Artificial Intelligence- A Review

Abstract

Precision plant breeding seeks to link biological target discovery with the development and evaluation of useful alleles in crop improvement. Genome-editing platforms, including CRISPR–Cas nucleases, base editors, prime editors, multiplex editing and cis-regulatory editing, enable targeted genomic alterations; however, their successful use depends on appropriate target choice, editing performance and phenotypic validation. Artificial intelligence (AI) offers complementary approaches for integrating pangenomic, multi-omics, phenotypic and environmental data to prioritise genes and regulatory sequences, predict guide-RNA activity and off-target effects, and support selection across genetic backgrounds and environments. This review examines the potential of combining these methods within a design–build–test–learn framework that connects edit design with molecular verification, high-throughput phenotyping, genomic prediction and conventional breeding operations. Examples discussed include susceptibility-gene and promoter editing for disease resistance, regulatory interventions associated with stress responses, nutritional modification and the emerging application of AI-designed nucleases in rice. Although these advances demonstrate technical feasibility, widespread implementation remains constrained by plant-specific data scarcity, differences among genotypes and environments, transformation and regeneration bottlenecks, complex trait architecture, unintended editing outcomes and regulatory diversity. The review emphasises prospective model validation, causal interpretation, pangenome-aware design, rigorous molecular characterisation and replicated field evaluation. It concludes that AI-guided genome editing can support crop improvement when used alongside quantitative genetics, breeder-defined objectives and responsible governance, rather than as a substitute for empirical breeding evidence.