singleCellHaystack

singleCellHaystack identifies differentially expressed genes (DEGs) in single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics data without requiring prior cell clustering, enabling detection of non-random gene expression patterns across cells.

Key Features:

• Clustering-independent analysis: Detects DEGs without relying on predefined cell clusters to avoid biases from cluster definitions.
• Utilization of Kullback-Leibler divergence: Uses Kullback-Leibler divergence to detect genes expressed in subsets of cells positioned non-randomly within multidimensional spaces.
• Compatibility with low-dimensional embeddings: Operates on data reduced by methods such as Principal Component Analysis (PCA) and on visualizations produced by t-SNE or UMAP.
• Demonstrated accuracy and efficiency: Comparative analyses on artificial datasets and application across 136 real transcriptome datasets and a spatial transcriptomics dataset indicate improved DEG prediction accuracy and practical efficiency.
• R package implementation: Provided as an R package for integration into computational workflows.

Scientific Applications:

• DEG discovery in scRNA-seq: Identification of genes with non-random expression patterns across single cells in scRNA-seq experiments.
• Analysis of cellular heterogeneity: Detection of subtle gene expression differences in heterogeneous populations such as during development, differentiation, or disease progression.
• Spatial transcriptomics analysis: Application to spatial transcriptomics data to find genes with spatially non-random expression patterns.

Methodology:

Performs clustering-independent detection of non-random gene expression using Kullback-Leibler divergence on gene expression distributions in multidimensional space, applied to reduced dimensions such as PCA and visualized with t-SNE or UMAP.