sc-REnF

sc-REnF applies entropy-based feature selection to single-cell RNA sequencing (scRNA-seq) data using Renyi and Tsallis entropies to identify informative genes for improved cell clustering and subtype resolution.

Key Features:

• Entropy-Based Gene Selection: Uses Renyi and Tsallis entropies to rank and select informative genes from scRNA-seq data.
• Robustness Against Technical Noise: Entropy-based selection reduces sensitivity to technical noise in scRNA-seq datasets.
• Improved Clustering Accuracy: Selected genes increase the purity and accuracy of cell clustering and classification.
• Application in Cell Annotation: Provides precise gene selections that enable classification and annotation of independent or unknown samples.

Scientific Applications:

• Single-Cell Typing: Refines cell-typing by providing informative and noise-resilient gene sets for downstream analyses.
• Subtype Detection: Enhances detection of subtle differences between cell subtypes by improving feature resolution.
• Marker Selection: Identifies robust marker genes for diagnostic and therapeutic studies based on single-cell data.

Methodology:

Uses Renyi and Tsallis entropies to select genes most informative for clustering, mitigating technical noise and enhancing the quality of gene selection.