UCell

UCell computes gene-signature scores for single-cell RNA sequencing data using the Mann-Whitney U statistic to provide robust scoring across heterogeneous and large datasets.

Key Features:

• Mann-Whitney U scoring: Generates signature scores using the Mann-Whitney U statistic.
• Cell-by-gene matrix input: Operates on any cell versus gene expression matrix.
• Seurat compatibility: Provides specialized functions to interact directly with Seurat objects.
• Scalability: Designed to process large-scale single-cell datasets.
• Robustness: Maintains scoring robustness against variations in dataset size and heterogeneity.
• Resource efficiency: Optimized for low time and memory requirements.
• Implementation: Available as an R package.

Scientific Applications:

• Gene-signature evaluation: Scoring and evaluation of gene expression signatures in single-cell RNA-seq data.
• Large-scale single-cell studies: Application to extensive single-cell datasets where computational efficiency is required.
• Comparative analysis across heterogeneous samples: Comparing signature activity across samples or conditions with varying cell composition and dataset sizes.

Methodology:

Computes signature scores by applying the Mann-Whitney U statistic to cell-by-gene expression matrices and exposes functions to operate directly on Seurat objects; implemented in R and optimized for low time and memory usage.