Mapping transcriptomic vector fields of single cells.

Area of Science:

• Computational Biology
• Systems Biology
• Genomics

Background:

• Single-cell RNA sequencing (scRNA-seq) provides high-resolution data on cellular states and transitions.
• Kinetic models are essential for understanding regulatory functions governing these cell dynamics.
• Existing RNA velocity analyses face limitations in accuracy and scope.

Purpose of the Study:

• Introduce Dynamo, an analytical framework for inferring RNA velocity and predicting cell fates.
• Develop a quantitative and predictive model for cell-state transitions.
• Overcome limitations of conventional splicing-based RNA velocity methods.

Main Methods:

• Inferred absolute RNA velocity using kinetic modeling.
• Reconstructed continuous vector fields to predict cell fates.
• Employed differential geometry to extract regulatory mechanisms.
• Applied least-action-path method for predicting transitions and perturbation outcomes.

Main Results:

• Accurate velocity estimations on metabolically labeled human hematopoiesis scRNA-seq data.
• Revealed mechanisms of megakaryocyte differentiation and PU.1-GATA1 circuit regulation.
• Successfully predicted drivers of hematopoietic transitions.
• In silico perturbations accurately predicted cell-fate diversions.

Conclusions:

• Dynamo advances quantitative and predictive theories of cell-state transitions.
• The framework enhances the analysis of scRNA-seq and RNA velocity data.
• Dynamo enables prediction of reprogramming paths and gene perturbation effects.