Extensible and developable genome prediction method

The invention relates to the crossing field of a biological information method and an artificial intelligence method, and relates to an extensible and developable genome prediction method, which comprises the following steps: obtaining a character phenotype of a target group and corresponding whole genome data to obtain a whole genome-phenotype data set; using a data processing component to process the whole genome-phenotype data set; dimension coding is carried out on the processed data, and adaptive coding is carried out after dimension coding is finished so as to meet the input requirement of a base model component required by base model construction; constructing a base model by utilizing a base model component based on the data subjected to adaptive coding; and converging and cutting the base model, and outputting a final prediction result. The prediction method provided by the invention solves the problems that more existing methods aim at a certain specific model, generalization and robustness in cross