How to define the internal layers related to informative for cell type, cell state, and common processes such as cell cycle?

[gancao]

[zhouzilu]

The short answer is that: we don't. Defining internal layers is a risky approach when I developed the methods, as it will introduce a lot more constraint and might be hard to transfer across cell type. I am not sure if there is any new development in the DL field but happy to know/discuss more.