FRMC

FRMC performs imputation of missing values in single-cell RNA sequencing (scRNA-seq) datasets using a singular value thresholding approximation to distinguish technical dropouts from true biological zeros.

Key Features:

• Fast and Accurate Imputation: The singular value thresholding approach improves computational speed while distinguishing technical dropouts from true biological zeros.
• Biological Relevance: Accurately imputes missing values attributed to technical noise, preserving gene expression profiles for downstream analysis of biological mechanisms.
• Enhanced Data Connectivity: Strengthens intracellular and intergenic connections within scRNA-seq data to support reconstruction of biological networks and pathways.
• Accurate Cell Clustering: Imputation improves separation and identification of distinct cell populations and functional states in clustering analyses.

Scientific Applications:

• Cellular Heterogeneity and Disease Mechanisms: Enables higher-resolution analysis of cellular heterogeneity and investigation of disease-related transcriptional changes.
• Large-Scale scRNA-seq of Complex Tissues: Applicable to large scRNA-seq datasets from complex tissues where missing values impede downstream inference.
• Rare Cell Types and Low RNA-Input Samples: Supports studies of rare cell populations and samples with limited RNA input by mitigating dropout effects.

Methodology:

Imputation is performed using a singular value thresholding approximation that differentiates technical dropouts from biological zeros and avoids reliance on pre-assumed data distributions.