scAnalysisToolkit

Overview

For internal use by the DeCarvalho and Constantinides labs at Scripps Research.

scAnalysisToolkit provides a comprehensive suite of helper functions for single-cell RNA-seq and ATAC-seq analysis workflows. The package includes helper tools and wrappers for:

• Clustering and Annotation: Seurat-based clustering workflows and cell type annotation with SingleR
• Differential Expression: Pseudobulk, Wilcoxon, and bulk RNA-seq differential expression analysis
• Differential Accessibility: ATAC-seq peak accessibility analysis
• Pathway Enrichment: GO term and Metascape enrichment analysis
• Repertoire Analysis: T-cell receptor repertoire analysis and visualization
• Cell Communication: Cell-cell communication analysis with CellChat
• Gene Programs: cNMF gene expression program analysis
• Motif Analysis: Motif enrichment in accessible chromatin regions
• Visualizations: Volcano plots, scatterplots, heatmaps, and circos diagrams

The package is designed to streamline end-to-end single-cell analysis with consistent interfaces and automated output generation, enabling iterative frictionless analysis of datasets. It builds on functionality provided by packages such as Seurat, Signac, SingleR, DESeq2, and CellChat.

Note: This package operates with Seurat objects in R only (no support for scran or scanpy).

Installation

You can install the development version of scAnalysisToolkit from GitHub with:

# install.packages("pak")
pak::pak("EduardAnsaldo/scAnalysisToolkit")

Usage

Section in construction…

The package provides high-level wrapper functions for common single-cell analysis workflows:

library(scAnalysisToolkit)

# Clustering workflow
clustering_results <- perform_seurat_clustering(
  seurat_object,
  dimensions = 30,
  resolutions = c(0.1, 0.25, 0.5)
)

# Cell type annotation
annotated_seurat <- annotate_seurat_with_SingleR_Eduard(
  seurat_object,
  local_path = "./results",
  database = "ImmGen"
)

# Differential expression analysis
de_results <- pseudobulk(
  seurat_object,
  comparison = "condition",
  group1 = "WT",
  group2 = "KO",
  cluster = "Tcells"
)