scdney

scdney performs differential cell-type composition analysis of single-cell RNA sequencing (scRNA-seq) data to estimate cell-type proportions and quantify their uncertainty across subjects and conditions.

Key Features:

• scDC method: Implements the single cell differential composition (scDC) statistical method to analyze cell-type compositions across subjects and conditions.
• Bootstrap resampling: Uses bootstrap resampling to estimate uncertainty in cell-type proportions and computes bias-corrected and accelerated bootstrap confidence intervals.
• Statistical modeling: Applies Generalized Linear Models (GLM) and Generalized Linear Mixed Models (GLMM) for significance testing of composition differences.
• Validation: Validated on simulated datasets and synthetic datasets derived from publicly available single-cell data, including scenarios with 2–5 subjects per condition.
• Implementation: Provided as an R package implementation for analysis of scRNA-seq compositional data.

Scientific Applications:

• Comparative composition analysis: Quantifying differences in cell-type composition across biological conditions or between individuals in scRNA-seq studies.
• Uncertainty quantification: Providing confidence intervals for cell-type proportions to support statistical comparisons of cellular composition.
• Cellular heterogeneity studies: Investigating cellular heterogeneity and its implications in biological processes and diseases using scRNA-seq data.

Methodology:

Implements the scDC statistical method; performs bootstrap resampling to generate bias-corrected and accelerated confidence intervals; applies GLM and GLMM for significance testing; validated on simulated and synthetic single-cell datasets.