Developmental scRNAseq Trajectories in Gene- and Cell-State Space-The Flatworm Example

Area of Science:

• Developmental Biology
• Computational Biology
• Genomics

Background:

• Single-cell RNA sequencing (scRNA-seq) captures cell transcriptomes for developmental studies.
• Computational methods infer pseudo-longitudinal differentiation trajectories from scRNA-seq data.
• Understanding cell adaptation to microenvironments requires analyzing transcriptional program dynamics.

Purpose of the Study:

• To introduce gene-state space trajectories as a complement to cell-state space trajectories.
• To visualize dynamic transcriptional programs during cell differentiation.
• To forecast RNA abundance changes using RNA velocities.

Main Methods:

• Utilized self-organizing maps (machine learning) to create 2D gene expression landscapes.
• Generated trajectories within these landscapes to represent developmental paths.
• Computed RNA velocities to create developmental vector fields for forecasting gene expression changes.

Main Results:

• Developed a method to generate gene-state space trajectories mirroring Waddington's epigenetic landscape.
• Visualized transcriptional programs dynamically changing during cell differentiation.
• Applied the method to planarian tissue development, illustrating its utility.

Conclusions:

• Gene-state space trajectories provide pseudo-temporal insights into changing transcriptional programs.
• This approach enhances the characterization of cell and tissue differentiation.
• Potential applications include studying aging, tumor progression, and integrating multi-omics data.