Subtypes estimation in single-cell multi-omics data via S-multi-SNE.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for transcriptomic analysis of individual cells.
• Analyzing scRNA-seq data presents challenges in visualizing and classifying cellular subtypes.
• Emerging multi-omics technologies enable simultaneous data acquisition from single cells.

Purpose of the Study:

• To adapt multi-view dimensionality reduction techniques (multi-SNE, S-multi-SNE) for scRNA-seq data analysis.
• To enhance the visualization and classification of cellular subtypes using scRNA-seq data.
• To integrate multi-omics datasets for improved cell subtype identification.

Main Methods:

• Adaptation of multi-SNE and S-multi-SNE for scRNA-seq data.
• Application of S-multi-SNE for cell subtype estimation.
• Integration of single-cell ATAC sequencing (scATAC-seq) and scRNA-seq data.

Main Results:

• S-multi-SNE successfully identified cell subtypes using data from the same or different species and sequencing technologies.
• S-multi-SNE demonstrated superior performance compared to other machine learning techniques.
• S-multi-SNE outperformed existing reference-based single-cell classification algorithms.
• Integration of scATAC-seq and scRNA-seq via S-multi-SNE improved cell subtype identification.

Conclusions:

• S-multi-SNE is a powerful tool for cell subtype identification in scRNA-seq data.
• The method is robust across diverse data types and species.
• Integrating multi-omics data with S-multi-SNE enhances biological insights.