scSampler

scSampler performs diversity-preserving subsampling of single-cell transcriptomic datasets to retain rare cell populations and representative cellular diversity for downstream analyses.

Key Features:

• Diversity-preserving subsampling: Subsamples single-cell transcriptomic data to retain representative cellular diversity, including low-abundance cell populations.
• Integration with Scanpy: Compatible with the Scanpy single-cell RNA-seq analysis pipeline for incorporation into Scanpy-based workflows.
• Python implementation: Implemented in Python for computational processing of large-scale single-cell datasets.
• R interface (rscsampler): Provides an R interface named rscsampler to enable use from R environments.

Scientific Applications:

• Exploratory data analysis: Facilitates fast exploration of cell type composition in large single-cell transcriptomic datasets by reducing dataset size while preserving diversity.
• Rare cell detection: Preserves low-abundance populations to support identification and characterization of rare cell types.
• Clustering and cell-type discovery: Produces representative subsamples that maintain cluster structure for downstream clustering and cell-type identification.
• Differential expression and trajectory inference: Enables downstream analyses such as differential expression testing and trajectory inference on reduced datasets that retain biological diversity.

Methodology:

Applies a subsampling algorithm that prioritizes retention of diverse and rare cell types to reduce dataset size while preserving representative cellular diversity.