Machine learning assisted breeding method and breeding chip

The invention discloses a machine learning assisted breeding method and a breeding chip, according to the breeding method, a specific phenotype of an organism is predicted mainly through specific tissue gene expression data in a specific development period under a specific growth condition of the organism, the breeding cost can be remarkably reduced, and the breeding efficiency and the success rate can be improved; the breeding chip is customized mainly based on the gene expression data, and can predict biological specific phenotypes through a machine learning optimal calculation model. The invention aims to solve the inherent defects of traditional machine learning assisted breeding based on genome data, such as difficulty in extracting related characteristic data of high-heterozygous and high-repetition complex genomes and no high-quality reference genome species and low accuracy. The invention provides a novel machine learning assisted breeding method for complex genome species, and the method is also suit