Trajectory Analysis of Tumor-Infiltrating Immune Cells

The status of the tumor-infiltrating immune cell subpopulations in the TME is dynamic and they may differentiate into different cellular states that exert different biological functions, e.g., cancer fighting or cancer tolerant. We performed the trajectory analysis using pseudo-time inferencing algorithm Monocle 2^{10 (link),50 (link)} to reconstruct the cell differentiation trajectory of different tumor-infiltrating immune cells. It uses a machine-learning technique called reversed graph embedding to describe multiple fate decisions in a fully unsupervised manner and derives a principal tree on a population of single cells that reveals the progression of cell and reconstruct their trajectory as a cell progresses through the biological process under study. Different branches in the cell trajectory likely distinguished molecularly distinct cell subpopulations (denoted by different cellular states) within a certain cell type.