Extracting Progressions in Biological Data by Identifying Branches in Potential of Heat-diffusion for Affinity-based Transition Embedding

As high dimensional biological data becomes more easily available, there is a pressing need for visualization tools that reveal the structure and emergent patterns of data in an intuitive form. Potential of Heat-diffusion for Affinity-based Transition Embedding (PHATE) [3] is a visualization method that captures both local and global nonlinear structure in data by an information-geometry distance between data points. The advantage of PHATE over many other visualization methods is its ability to reveal branching structures that commonly exist in differentiation systems. For example, PHATE can faithfully visualize the underlying trajectories in a newly generated scRNA-seq dataset of human germ layer differentiation. Here, PHATE reveals a dynamic picture of the main developmental branches in unparalleled detail.