GeneClust

GeneClust selects feature genes for single-cell RNA sequencing (scRNA-seq) cell clustering by optimizing gene relevance while minimizing redundancy and preserving complementarity.

Key Features:

• Relevance Maximization: Evaluates gene expression profiles to prioritize genes that are highly informative for distinct cell types in scRNA-seq datasets.
• Redundancy Minimization: Groups genes with similar expression patterns to reduce redundant features and improve clustering resolution.
• Complementarity Preservation: Preserves complementary gene interactions to capture diverse biological processes within datasets.
• Integration Capability: Designed as a plug-in, it integrates with existing cell clustering methods to supply selected feature genes for downstream clustering analyses.

Scientific Applications:

• Enhanced Clustering Performance: Extensive benchmarking demonstrates significant improvements in the accuracy and reliability of scRNA-seq cell clustering when using GeneClust-selected features.
• Investigation of Gene Interactions: Groups cofunctional genes involved in biological processes and pathways to facilitate study of gene interactions within specific cellular contexts.
• Identification of Biologically Relevant Genes: Aids identification of genes pertinent to dataset-specific biological characteristics, supporting insight into cellular functions and disease mechanisms.

Methodology:

Groups genes by expression profiles, then selects a subset that maximizes relevance to cell types while minimizing redundancy and preserving complementarity among gene clusters.