cell2fate

cell2fate improves RNA velocity analysis by decomposing realistic biophysical models of transcription dynamics, providing a biophysical connection between RNA velocity and statistical dimensionality reduction.

Key Features:

• Linearized velocity ODE formulation: cell2fate formulates RNA velocity based on a linearization of the velocity ordinary differential equation, enabling a fully Bayesian solution of a biophysically more accurate model.
• RNA velocity module decomposition: cell2fate decomposes RNA velocity solutions into modules, establishing a biophysical connection between RNA velocity and statistical dimensionality reduction.
• Enhanced detection of weak dynamical signals: cell2fate demonstrates improved power to reconstruct complex dynamics and weak dynamical signals in rare and mature cell types.
• Spatial mapping of RNA velocity modules: cell2fate enables spatial mapping of RNA velocity modules onto tissue architecture, connecting spatial organization with temporal transcriptional dynamics.

Scientific Applications:

• Cell fate prediction: cell2fate infers RNA velocity modules to improve prediction of cell fate from single-cell RNA sequencing data.
• Reconstruction of complex transcriptional dynamics: cell2fate reconstructs complex or weak transcription rate changes across cellular trajectories that challenge existing velocity models.
• Spatial transcriptomics analysis: cell2fate maps RNA velocity modules onto tissue architecture, as demonstrated in the developing human brain.

Methodology:

cell2fate employs a linearization of the velocity ordinary differential equation to solve a biophysically realistic model of transcription dynamics in a fully Bayesian fashion, decomposing RNA velocity solutions into modules.