Cell cycle gene regulation dynamics revealed by RNA velocity and deep-learning

Area of Science:

• Molecular Biology
• Genomics
• Computational Biology

Background:

• The cell cycle is crucial for life, but its gene regulation dynamics are not fully understood.
• Single-cell RNA sequencing (scRNA-seq) offers a way to study these dynamics non-invasively.

Purpose of the Study:

• To develop an efficient computational method for analyzing cell cycle gene dynamics from scRNA-seq data.
• To create a high-resolution map of the cell cycle transcriptome.

Main Methods:

• Generating scRNA-seq libraries across various cell systems.
• Observing cycling patterns in unspliced-spliced RNA.
• Applying a novel deep learning approach, DeepCycle, to model these patterns.

Main Results:

• Identified cyclical patterns in cell cycle-related gene expression using scRNA-seq.
• Developed DeepCycle to accurately map the cell cycle transcriptome at high resolution.
• Characterized major transcriptional waves during the G1 phase in embryonic and somatic cells.

Conclusions:

• DeepCycle provides an effective tool for analyzing cell cycle dynamics from scRNA-seq data.
• The study offers a detailed map of the cell cycle transcriptome, advancing our understanding of cell division.
• This work facilitates future studies of the cell cycle across diverse biological contexts.