ILoReg

ILoReg applies probabilistic feature extraction to single-cell RNA sequencing (scRNA-seq) data to improve resolution of cell population identification and detect rare subpopulations, and is implemented as an R package.

Key Features:

• Probabilistic Feature Extraction: Performs a probabilistic feature extraction step that enhances differentiation resolution among cells by reducing dimensionality while preserving critical transcriptomic information.
• Unsupervised Clustering and Visualization: Uses the extracted features for unsupervised clustering of cells and provides visualization of identified subpopulations.
• Identification of Rare Cell Populations: Enables detection of rare cell populations that are often missed by conventional scRNA-seq analysis pipelines.

Scientific Applications:

• High-resolution cell type and state identification: Resolves subtle transcriptomic differences to distinguish closely related cell populations and states within complex tissues.
• Oncology / Cancer research: Facilitates discovery of tumor subpopulations and heterogeneity, including rare malignant or microenvironmental cell types.
• Immunology: Supports identification of diverse immune cell subsets and rare immune states in single-cell immune profiling.
• Developmental biology and stem cell research: Detects transient or rare developmental cell states and stem cell subpopulations during differentiation processes.

Methodology:

Applies a probabilistic feature extraction step prior to clustering and visualization to reduce data dimensionality while preserving critical information, followed by unsupervised clustering algorithms and visualization techniques.