ZINBMM: a general mixture model for simultaneous clustering and gene selection using single-cell transcriptomic data
Yang Li1,2,3, Mingcong Wu1,3, Shuangge Ma4
1Center for Applied Statistics and School of Statistics, Renmin University of China, Beijing, China.
Genome Biology
|September 11, 2023
Summary
This study introduces a novel zero-inflated negative binomial mixture model (ZINBMM) for single-cell RNA sequencing (scRNA-seq) data analysis. ZINBMM effectively clusters cells and identifies cluster-specific genes, enhancing the understanding of cell heterogeneity.
Area of Science:
- Genomics
- Computational Biology
- Bioinformatics
Background:
- Single-cell RNA sequencing (scRNA-seq) is crucial for identifying cell types and lineages.
- Existing methods often lack the ability to pinpoint cluster-specific genes driving cell heterogeneity.
- Understanding cell heterogeneity is key to advancing biological insights.
Purpose of the Study:
- To develop a novel computational model for simultaneous clustering and gene selection in scRNA-seq data.
- To address limitations in current methods for investigating cluster-specific genes.
- To improve the biological understanding of cell heterogeneity.
Main Methods:
- A zero-inflated negative binomial mixture model (ZINBMM) was developed.
- The model analyzes raw gene expression counts, accounting for batch effects and dropout events.
- Systemic analysis was performed on simulated and real scRNA-seq datasets.
Main Results:
- ZINBMM demonstrated effective clustering of scRNA-seq data.
- The model successfully performed gene selection for cluster-specific markers.
- Practical applicability was validated across five diverse scRNA-seq datasets.
Conclusions:
- ZINBMM offers a robust approach for scRNA-seq data analysis.
- The method enhances the identification of genes contributing to cell heterogeneity.
- ZINBMM facilitates deeper biological insights from single-cell data.
