# Bayesian inference of RNA velocity incorporating timepoints, lineage bifurcations, and count data

## TL;DR

VeloVAE is a new Bayesian method that improves the accuracy of tracking gene expression changes in cells over time, even when data is limited to single snapshots.

## Contribution

VeloVAE introduces a Bayesian framework for RNA velocity that models lineage bifurcations, discrete count data, and latent time with improved accuracy.

## Key findings

- VeloVAE outperforms existing methods in reconstructing gene expression dynamics and estimating transcription rates.
- The model accurately captures complex biological systems like hematopoiesis and brain development.
- Automatically inferred latent time performs better than manually defined pseudotime in some cases.

## Abstract

Experimental approaches for measuring single-cell gene expression can observe each cell at only one time point, requiring computational approaches for reconstructing the dynamics of gene expression during cell fate transitions. RNA velocity is a promising computational approach for this problem, but existing inference methods fail to capture key aspects of real data, limiting their utility. To address these limitations, we developed VeloVAE, a Bayesian model for RNA velocity inference. VeloVAE uses variational Bayesian inference to estimate the posterior distribution of latent time, latent cell state, and kinetic rate parameters for each cell. Our approach can incorporate prior distributions on rate parameters and time points; model lineage bifurcations using branching differential equations; and directly model discrete count data. We show that VeloVAE significantly outperforms previous approaches in terms of data fit, accuracy of inferred differentiation directions, and transcription rate estimation. These improvements allow VeloVAE to accurately model gene expression dynamics in complex biological systems, including hematopoiesis, induced pluripotent stem cell reprogramming, the developing mouse brain, and the entire mouse embryo. We find that the latent time automatically inferred using all cells can even outperform pseudotime inferred using manually chosen cell subsets and root cells. Our work provides important new tools for modeling sequential changes in gene expression from single-cell expression data.

Understanding how cells change their identity over time is a central question in biology. However, most experiments that measure gene activity in individual cells can only capture a single snapshot of each cell, making it difficult to see how cells move through developmental or disease-related processes. In our work, we developed a computational method VeloVAE to reconstruct these cellular trajectories from scRNA-seq data. Our approach uses a Bayesian framework to infer the latent time, cell state and the rates at which genes turn on or off. By integrating biological information and modeling complex processes such as cell type bifurcation, VeloVAE provides a more accurate and interpretable view of cell development than previous approaches. We demonstrate that VeloVAE can reveal key patterns of gene activity across several biological systems, including blood formation, brain development, and stem cell reprogramming. Overall, our work provides a general framework for understanding dynamic gene regulation from single-cell data.