Multi-Objective Optimized Fuzzy Clustering for Detecting Cell Clusters from Single-Cell Expression Profiles.

Area of Science:

• Genomics and Bioinformatics
• Computational Biology
• Molecular Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) enables gene expression analysis at single-cell resolution.
• Accurate cell cluster detection from scRNA-seq data remains a significant challenge in biological research.

Purpose of the Study:

• To develop and validate a multi-objective optimization-based fuzzy clustering approach for scRNA-seq data analysis.
• To identify cell clusters and differentially expressed genes (DEGs) for biological discovery.

Main Methods:

• Data preprocessing including filtering and SCnorm normalization.
• Application of fuzzy c-means clustering with evaluation using Partition Entropy (PE), Partition Coefficient (PC), Modified Partition Coefficient (MPC), and Fuzzy Silhouette Index (FSI).
• Multi-objective decision-making using TOPSIS to determine optimal clustering, followed by Limma for DEG analysis.

Main Results:

• The proposed method successfully clustered a mouse intestinal cell scRNA-seq dataset (GSE62270) into two distinct cell populations.
• Identified 1240 differentially expressed genes between the two clusters.
• Discovered Khk (ketohexokinase) as a novel marker for a rare intestinal cell type.

Conclusions:

• The multi-objective fuzzy clustering approach provides a robust method for cell cluster detection in scRNA-seq data.
• This technique aids in identifying novel biomarkers for rare cell populations.
• The method enhances the interpretability and discovery potential of scRNA-seq datasets.