FLASH-MM: Fast and Scalable Single-Cell Differential Expression Analysis Using Linear Mixed-Effects Models

FLASH-MM is a fast and scalable algorithm for differential expression (DE) analysis in large-scale single-cell RNA-seq (scRNA-seq) datasets. It addresses challenges such as intra-subject correlation, inter-subject variability, and the computational demands of analyzing millions of cells.

Key Features

• Efficient and Scalable: Precomputes summary statistics to handle large datasets while maintaining single-cell resolution.
• Accurate DE Analysis: Controls type-I error rates and maintains high statistical power.
• Broad Applications: Supports case-control comparisons, cell-type-specific analyses, and multi-subject studies.
• Simulation Tool: Includes simuRNAseq for generating realistic scRNA-seq datasets.

Applications

FLASH-MM has been applied to:

• Case-control comparisons in tuberculosis immune atlases.
• Cell-type-specific sex comparisons in kidney datasets.

With its speed, accuracy, and flexibility, FLASH-MM enables robust DE analysis for large-scale single-cell studies across diverse biological contexts.