Machine Learning for AI Breeding in Plants

What makes artificial intelligence (AI) smart is machine learning (ML), a "field of study that gives computers the ability to learn without being explicitly programmed", as defined by ML pioneer Arthur Samuel in 1959.ML deduces data patterns without relying on prior assumptions as statistics does, greatly reducing the human effort required to understand the data.ML comprises a large family of algorithms, many of which support big data analytics [1].With the rapid advances in multi-omics technologies, plant breeding has entered the "genome, germplasm, genes, genomic breeding, and gene editing (5G)" generation [2], in which biological knowledge and omics data are integrated to expedite trait improvement.ML holds great promise for 5G breeding, with many reports of ML applications for omics-driven gene discovery, genotype-to-phenotype (G2P) prediction, genomic selection (GS), and plant phenomics.However, there remains a gap between basic research and breeding practices in plants [3].Given multi-omics, genotypic, phenomic, and environmental datasets have become highly dimensional and heterogeneous, novel ML algorithms are expected.Hereby, we propose ways to overcome major challenges in the application of cutting-edge ML models to plant research, with the ultimate goal of making plant breeding smart and easy.