| HRA005590
(Open Access)
|
Deciphering TF activity is essential for gaining a better understanding of the uniqueness and functionality of each cell type. Herein, we introduce metaTF (https://github.com/wanglabsmu/metaTF), a computational machine learning framework designed to infer TF activity in scRNA-seq data, and outperforms existing methods in estimating TF activity. It presents the improved effectiveness in characterizing cell identity during mouse hematopoietic stem cell development. Furthermore, metaTF provides a superior characterization of the functional identity of breast cancer epithelial cells, and newly identifies a subset of neural-regulated T cells within the tumor immune microenvironment, which potentially activates BCL6 in response to neural-related signals. Overall, metaTF enables robust TF activity analysis from scRNA-seq data, significantly enhancing characterization of cell identity and function. |
