scMC learns biological variation through the alignment of multiple single-cell genomics datasets

Summary

scMC is a new method for integrating single-cell genomics datasets that effectively removes technical variation while preserving biological variation, enabling accurate alignment and detection of biological signals.

Key findings

• scMC successfully removes technical variation.
• scMC preserves biological variation.
• scMC accurately aligns simulated and real single-cell datasets.
• scMC detects context-shared and context-specific biological signals.

Methods

• scMC integration method for single-cell genomics data.
• Unsupervised inference of technical variation.
• Variance analysis to learn biological variation.
• Alignment of multiple single-cell genomics datasets.

Limitations

• Existing methods struggle to distinguish biological from technical variation.
• Existing methods may remove both biological and technical variation.

Applications

• Integration of single-cell RNA-seq datasets.
• Integration of single-cell ATAC-seq datasets.
• Analysis of developmental biology data.
• Analysis of cancer research data.