scBridge embraces cell heterogeneity in single-cell RNA-seq and ATAC-seq data integration

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell multi-omics data integration seeks to align different data types while preserving cell identity.
• Cellular heterogeneity complicates distinguishing between omics and cell-type specific variations.
• Existing methods often struggle to account for this inherent biological variability.

Purpose of the Study:

• To develop a novel method for single-cell multi-omics data integration that leverages cell heterogeneity.
• To improve the accuracy and robustness of integrating diverse omics datasets from single cells.
• To address the challenge of distinguishing omics differences from cell-type differences in heterogeneous cell populations.

Main Methods:

• Proposed scBridge, a heterogeneous multi-omics integration method.
• scBridge iteratively identifies cells with minimal omics differences (reliable cells).
• Integrates these reliable cells with other omics data (e.g., scRNA-seq) to bridge the omics gap.

Main Results:

• Demonstrated the effectiveness of exploiting cell heterogeneity for data integration.
• scBridge successfully reduced omics differences while maintaining cell type distinctions.
• Outperformed six representative baseline methods across seven multi-omics datasets.

Conclusions:

• Cell heterogeneity is a valuable feature, not just noise, for multi-omics integration.
• scBridge offers a superior approach to single-cell multi-omics data integration by handling heterogeneity.
• The method enhances the ability to analyze complex biological systems using multi-omics data.