Accurate Transcription Factor Activity Inference to Decipher Cell Identity from Single-Cell Transcriptomic Data with MetaTF.

Abstract

Cellular heterogeneity within cancer tissues determines cancer progression and treatment response. Single-cell RNA sequencing (scRNA-seq) has provided a powerful approach for investigating the cellular heterogeneity of both cancer cells and stroma cells in the tumor microenvironment. However, the common practice to characterize cell identity based on the similarity of their gene expression profiles may not really indicate distinct cellular populations with unique roles. Generally, the cell identity and function are orchestrated by the expression of given specific genes tightly regulated by transcription factors (TFs). Therefore, deciphering TF activity is essential for gaining a better understanding of the uniqueness and functionality of each cell type. Herein, metaTF, a computational framework designed to infer TF activity in scRNA-seq data, is introduced and existing methods are outperformed for 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 identifies a novel 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 the characterization of cell identity and function.