scRNASequest

scRNASequest processes single-cell RNA sequencing (scRNA-seq) data to provide preprocessing, harmonization, reference-based cell type label transfer and embedding projection, multi-sample multi-condition differential gene expression analysis, and export for visualization and data sharing.

Key Features:

• Preprocessing: Preprocesses raw Unique Molecular Identifier (UMI) count data for downstream analysis.
• Harmonization: Performs dataset harmonization using one or multiple methods to integrate data across experimental conditions or platforms.
• Cell Type Label Transfer and Embedding Projection: Supports reference-dataset-based cell type label transfer and projection of embeddings onto reference spaces.
• Differential Gene Expression Analysis: Conducts multi-sample, multi-condition differential gene expression analysis at the single-cell level.
• Integration with Visualization and Data Sharing Tools: Produces outputs compatible with cellxgene VIP and CellDepot for visualization and hosting.
• Output Formats: Generates h5ad files for interoperability with downstream tools.

Scientific Applications:

• Hypothesis generation and discovery: Identifies gene expression patterns and candidate markers from scRNA-seq data to support hypothesis generation.
• Comparative analyses across conditions: Enables comparative analysis of cell-type-specific expression across multiple samples and conditions using harmonization and multi-condition DGE.
• Cell type annotation: Facilitates annotation of cell types through reference-based label transfer and embedding projection.
• Domain-specific research: Supports applications in developmental biology, immunology, and cancer research by enabling standardized single-cell analyses.

Methodology:

Preprocessing of raw UMI count data; dataset harmonization using one or multiple methods; reference-dataset-based cell type label transfer and embedding projection; multi-sample, multi-condition differential gene expression analysis at the single-cell level; generation of h5ad output files; semi-automated workflow with customizable steps; execution on Linux/Unix (including MacOS) or on HPC clusters using SGE/Slurm.