Clustering of single-cell multi-omics data with a multimodal deep learning method.

Area of Science:

• Single-cell biology
• Computational biology
• Genomics

Background:

• Single-cell multimodal sequencing enables simultaneous profiling of diverse molecular data within individual cells.
• Accurate cell type identification from multimodal data is crucial for downstream biological analyses.
• Integrating multiple data sources for single-cell clustering presents significant statistical and computational challenges.

Purpose of the Study:

• To introduce scMDC, a novel deep learning method for single-cell multi-omics data clustering.
• To address the challenge of jointly analyzing diverse single-cell data modalities for improved clustering.
• To provide a scalable and effective computational tool for multimodal single-cell data analysis.

Main Methods:

• scMDC employs an end-to-end deep learning architecture.
• The model explicitly characterizes different data modalities.
• It jointly learns latent features for integrated clustering analysis.

Main Results:

• scMDC demonstrates superior performance compared to existing single-modal and multimodal clustering methods.
• The method shows strong results on both simulated and real single-cell multimodal datasets.
• scMDC exhibits linear scalability, making it suitable for large-scale datasets.

Conclusions:

• scMDC offers a robust solution for single-cell multi-omics data clustering.
• The method enhances the accuracy of cell type identification by effectively integrating multimodal information.
• Its scalability and performance position scMDC as a valuable tool for future single-cell research.