Non-Negative Low-Rank Representation With Similarity Correction for Cell Type Identification in scRNA-Seq Data

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-Seq) is crucial for understanding cellular heterogeneity.
• Identifying cell types is a key step in scRNA-Seq data analysis.
• Existing methods often neglect intercellular structural relationships.

Purpose of the Study:

• To develop a novel model for cell type identification that preserves both global and local cellular structures.
• To improve the accuracy of cell type identification in scRNA-Seq data.

Main Methods:

• Introduced a non-negative low-rank similarity correction (NLRSIM) model.
• Utilized subspace clustering to maintain global cell structure.
• Incorporated manifold learning and a position-sensitive hashing algorithm to preserve local geometric structures.

Main Results:

• NLRSIM demonstrated superior clustering performance compared to existing advanced models.
• Visualization experiments confirmed the effectiveness of NLRSIM.
• Validated gene expression information calibrated by NLRSIM in biological studies.

Conclusions:

• NLRSIM offers a novel approach to cell type identification by preserving intercellular structural relationships.
• The model provides deeper insights into gene expression, cellular states, and structures.
• NLRSIM contributes new perspectives to single-cell data analysis and biological discovery.