scverse/cellrank

What it solves

CellRank addresses the challenge of mapping cellular dynamics and fate mapping in single-cell data. It allows researchers to determine how cells differentiate and which paths they take, moving beyond simple trajectory inference to provide probabilistic estimates of cell fates.

How it works

The framework uses Markov state modeling of multi-view single-cell data to study cellular dynamics. It integrates various biological priors—such as RNA velocity, pseudotime, developmental potential, or experimental time points—to estimate the direction of differentiation and compute the probabilities of cells reaching specific terminal or intermediate states.

Who it’s for

It is designed for biologists and bioinformaticians working with single-cell analysis, specifically those looking to model cell differentiation, identify driver genes, and analyze gene expression trends across cellular transitions.

Highlights

• Scalable to large numbers of cells.
• Fully compatible with the scverse ecosystem.
• Ability to infer fate probabilities and identify driver genes.
• Supports multiple biological priors for estimating differentiation direction.