Quantifying uncertainty in RNA velocity

The concept of RNA velocity has made it possible to extract dynamic information from single-cell RNA sequencing data snapshots, attracting considerable attention and inspiring various extensions. Nonetheless, existing approaches lack uncertainty quantification and many adopt unrealistic assumptions or employ complex black-box models that are difficult to interpret. In this paper, we present a Bayesian hierarchical model to estimate RNA velocity, which leverages a time-dependent transcription rate and non-trivial initial conditions, allowing for well-calibrated uncertainty quantification. The proposed method is validated in a comprehensive simulation study that covers various scenarios, and benchmarked against a widely embraced and commonly recognized approach for RNA velocity on single-cell RNA sequencing data from mouse embryonic stem cells. We demonstrate that our model surpasses this widely used, state-of-the-art method, offering enhanced interpretation of cell velocity and cell orders. Additionally, it supports the estimation of a unified gene-shared latent time, providing a valuable resource for downstream analysis.