SCSMD: Single Cell Consistent Clustering based on Spectral Matrix Decomposition.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) analysis relies on cluster analysis to understand cellular heterogeneity.
• Existing clustering algorithms often produce inconsistent results in terms of cluster number and assignments.
• Accurate cell type identification and understanding of developmental trajectories are crucial in scRNA-seq.

Purpose of the Study:

• To introduce a novel, comprehensive clustering approach, Single Cell Consistent Clustering based on Spectral Matrix Decomposition (SCSMD).
• To enhance the accuracy and consistency of cluster identification in scRNA-seq data.
• To validate SCSMD's performance across various datasets and cell types.

Main Methods:

• Developed Single Cell Consistent Clustering based on Spectral Matrix Decomposition (SCSMD).
• Integrated multiple clustering methods to determine an optimal clustering scheme.
• Validated SCSMD using 15 authentic scRNA-seq datasets and a bespoke evaluation metric.

Main Results:

• SCSMD demonstrated robust performance in cluster number and assignment accuracy across diverse scRNA-seq datasets.
• Applied to human embryonic stem cells, SCSMD identified known cell types and developmental trajectories.
• In glioblastoma cells, SCSMD accurately detected cell types and provided finer sub-clustering within existing clusters.

Conclusions:

• SCSMD offers a superior and consistent approach to cluster analysis in scRNA-seq data.
• The method is effective for identifying cell types, developmental trajectories, and finer cellular distinctions.
• SCSMD is scalable to larger datasets and suitable for further downstream analyses in scRNA-seq research.