SLRRSC: Single-Cell Type Recognition Method Based on Similarity and Graph Regularization Constraints.

Area of Science:

• Computational Biology
• Genomics
• Data Science

Background:

• Single-cell RNA sequencing (scRNA-seq) enables cell type identification.
• High noise and dimensionality in scRNA-seq data present challenges for accurate cell clustering.
• Existing methods struggle with effectively identifying cell types from complex cell mixtures.

Purpose of the Study:

• To propose a novel subspace clustering algorithm, SLRRSC, for improved scRNA-seq data analysis.
• To enhance cell type recognition by accurately capturing global and local data properties.
• To address the limitations of current methods in handling noisy and high-dimensional scRNA-seq data.

Main Methods:

• Developed SLRRSC based on the low-rank representation (LRR) model.
• Incorporated manifold-based graph regularization to preserve local geometric data structures.
• Introduced a similarity constraint to enhance the learning of global data structures and ensure matrix symmetry.

Main Results:

• SLRRSC demonstrated superior performance compared to other single-cell clustering methods.
• The algorithm generated a more accurate sample similarity matrix.
• Achieved effective cell type recognition on both simulated and real scRNA-seq datasets.

Conclusions:

• SLRRSC offers an effective approach for single-cell clustering and cell type identification.
• The integration of graph regularization and similarity constraints improves the accuracy and interpretability of LRR-based methods.
• SLRRSC advances scRNA-seq data analysis by providing a robust solution for noisy, high-dimensional biological data.