scEFSC

scEFSC performs ensemble feature selection and consensus clustering to improve accuracy and biological interpretability of single-cell RNA-seq (scRNA-seq) data clustering.

Key Features:

• Systematic Performance Evaluation: Evaluates four feature selection methods and nine clustering algorithms across fourteen real scRNA-seq datasets to identify strategies for high-dimensional data, numerical instability, and computational scalability.
• Ensemble Feature Selection: Employs multiple unsupervised feature selection techniques to filter genes that do not contribute to clustering, increasing robustness of selected features.
• Consensus Clustering: Applies various scRNA-seq clustering algorithms to the filtered data and integrates results using a weighted-based meta-clustering strategy to mitigate individual algorithm biases.
• Performance Superiority: Demonstrates improved clustering performance relative to existing algorithms based on benchmarking across fourteen real datasets and multiple evaluation metrics.
• Biological Interpretability: Conducts differential gene expression analysis, gene ontology enrichment, and KEGG pathway analysis to validate the biological relevance of identified clusters.

Scientific Applications:

• Developmental Biology: Resolves cellular heterogeneity and developmental trajectories at single-cell resolution.
• Cancer Heterogeneity: Identifies tumor subpopulations and transcriptional programs underlying intratumoral diversity.
• Immune Response Studies: Characterizes diverse immune cell states and responses within complex tissues.
• General Cellular Diversity Analysis: Profiles diverse cell populations in studies requiring detailed transcriptomic clustering and interpretation.

Methodology:

Evaluates four feature selection methods and nine clustering algorithms across fourteen real scRNA-seq datasets; applies multiple unsupervised feature selection techniques to filter genes; runs various scRNA-seq clustering algorithms and integrates results via a weighted-based meta-clustering strategy; performs differential gene expression analysis, gene ontology enrichment, and KEGG pathway analysis.