distinct

distinct performs differential analysis of full distributions to detect complex expression differences between conditions in single-cell RNA sequencing (scRNA-seq) and high-dimensional flow or mass cytometry (HDCyto) data.

Key Features:

• Distribution-level analysis: Compares full distributions rather than focusing solely on changes in the mean.
• Supported data types: Applicable to single-cell RNA sequencing (scRNA-seq) and high-dimensional flow or mass cytometry (HDCyto) datasets.
• Detection of non-mean differences: Identifies both prominent mean shifts and subtle differential patterns that do not affect the mean.
• Statistical framework: Employs a hierarchical non-parametric permutation approach for inference.
• ECDF comparison: Detects distributional differences by comparing empirical cumulative distribution functions (ECDFs) across conditions.
• Benchmark performance: Demonstrated improved sensitivity for identifying a greater number of differential patterns on simulated and experimental scRNA-seq and mass cytometry datasets.
• Error-rate control: Maintains control over false positive rate and false discovery rate in evaluations.
• Implementation: Provided as an R package implementation.

Scientific Applications:

• Single-cell differential expression: Detects complex expression changes in cell populations from scRNA-seq beyond mean-based effects.
• Cytometry differential analysis: Identifies distributional differences in high-dimensional flow and mass cytometry (HDCyto) measurements between conditions.
• Cellular heterogeneity characterization: Reveals subtle shifts in subpopulation distributions to inform studies of cellular heterogeneity.

Methodology:

Uses a hierarchical non-parametric permutation approach that compares empirical cumulative distribution functions (ECDFs) across conditions.