MomicPred: A Cell Cycle Prediction Framework Based on Dual-Branch Multi-Modal Feature Fusion for Single-Cell...

Area of Science:

• Cell Biology
• Genomics
• Bioinformatics

Background:

• The cell cycle is critical for cell fate and stem cell differentiation, impacting tissue homeostasis.
• Accurate cell cycle regulation is essential for maintaining cellular diversity.
• Single-cell multi-omics technologies offer potential for improved cell cycle prediction by integrating gene expression and chromatin data.

Purpose of the Study:

• To develop an advanced computational framework for predicting cell cycle dynamics.
• To effectively integrate transcriptome and 3D genome data for enhanced cell cycle prediction.
• To uncover cross-layer associations between gene expression and chromatin structure in cell cycle regulation.

Main Methods:

• Proposed MomicPred, a dual-branch multi-modal fusion framework.
• Integrated transcriptome-derived gene expression data with 3D genome structural insights.
• Extracted three core feature sets to capture synergistic interactions between omics modalities.

Main Results:

• Achieved high-precision cell cycle prediction by leveraging complementary multi-omics data.
• Demonstrated the efficiency and robustness of MomicPred through benchmarking.
• Identified key biological processes and chromatin structural changes across distinct cell cycle stages.

Conclusions:

• MomicPred offers a novel approach for predicting cell cycle dynamics using integrated multi-omics data.
• The framework provides new perspectives on the interplay between gene expression and chromatin structure.
• Findings contribute to a deeper understanding of cell cycle regulation in stem cell differentiation and tissue homeostasis.