Single cell type annotation model based on constrained domain adaptive network

The invention discloses a single cell type annotation model scCDAN based on a constrained domain adaptive network. According to the invention, by using the domain alignment module, data distribution alignment of the source domain and the target domain is realized through an adversarial learning strategy, so that inter-domain difference is reduced. In order to better carry out fine-grained distinguishing on cell types, a type boundary constraint module is designed, the position relation between cells of the same type and cells of different types in a feature space is constrained through ternary loss and center loss, and the distinguishing performance between the cells is improved. And respectively carrying out experiments on the simulation data set, the cross-platform data set and the cross-species data set to verify the effectiveness of the scCDAN model. Experimental results show that the scCDAN model has good performance in cell type annotation and batch correction, and is beneficial for helping researchers