scVelo

scVelo models RNA velocity from single-cell RNA-sequencing by estimating transcriptional dynamics from spliced and unspliced mRNA to characterize cellular differentiation and transient cell states.

Key Features:

• Likelihood-based dynamical modeling: Employs a likelihood-based dynamical model that solves the full splicing kinetics to generalize RNA velocity to transient cell states.
• Gene-specific kinetic inference: Infers gene-specific rates of transcription, splicing, and degradation.
• RNA velocity estimation from spliced and unspliced mRNA: Estimates rates of gene expression change by analyzing spliced and unspliced mRNA counts.
• Cell positioning along differentiation pathways: Determines cell positions along differentiation trajectories using inferred transcriptional dynamics.
• Identification of putative driver genes: Detects putative driver genes that may influence cell fate decisions based on inferred kinetics and velocity vectors.

Scientific Applications:

• Neurogenesis: Applied to neurogenesis to disentangle subpopulation kinetics and resolve dynamic developmental trajectories.
• Pancreatic endocrinogenesis: Applied to pancreatic endocrinogenesis to resolve transient endocrine progenitor states and lineage emergence.
• Cellular differentiation, lineage decisions, and gene regulation: Used to study differentiation processes, infer lineage decisions, and investigate gene regulatory dynamics in transient and non–steady-state systems.

Methodology:

Uses a likelihood-based dynamical model to solve full splicing kinetics, infers gene-specific transcription, splicing, and degradation rates, and estimates RNA velocity from spliced and unspliced mRNA counts.