This document discusses using super-enhancer RNA profiles to identify cell states. The presenter hypothesizes that each cell state can be represented by a distinct super-enhancer RNA profile. They analyze a dataset containing various cell types and conditions to construct super-enhancer RNA profiles. Non-negative matrix factorization is used to decompose the data and reveal hidden cell states. Their results suggest super-enhancer RNA profiles can effectively classify cell types and identify state transitions, such as during induced pluripotent stem cell differentiation. Future work involves using these methods to study core regulatory circuits and cell state changes, especially related to tumorigenesis.