Trajectory-Based Functional Clustering of Single Cell RNA Sequencing Data

Abstract

As a high-throughput sequencing technique, single cell RNA sequencing (scRNA-seq) allows for the measurement of gene expression in a single cell. This cell-specific transcriptome landscape enables researchers to reveal cellular developmental sequences such as growth and differentiation of single cells using pseudotime estimation. Identifying the genes associated with such developmental processes is critical to understand these biological processes. In this project, we propose a novel trajectory-based functional clustering algorithm to cluster genes over pseudotime points. In the algorithm, we combine the expectation–maximization (EM) algorithm with nonparametric mixed-effect models to estimate the cluster labels as well as model the pseudotemporal dependency of gene expression levels over pseudotime points.