scooter

Exploratory single-cell data analysis at the sample level

scxplor is designed to help you explore your single-cell RNA-seq data in a simple and time-efficient way. It summarizes your cell type annotations by providing tools for compositional data analysis as well as tools for gene expression on the sample and cell type level.

• Supervised analysis: visualize your samples with box plots comparing groups or PCA colored by group
• Unsupervised analysis: cluster your samples based on their similarity

Why unsupervised analysis?

• Quality Control: Identify outliers, biases, and potential technical artifacts.
• Dimensionality Reduction: Condense 1000s of dimensions into a few highly interpretable features.
• Detection of Biological Variability: reveal important insights into population heterogeneity, developmental trajectories, and responses to stimuli or disease.

Installation

# install.packages("remotes")
remotes::install_github("carmonalab/scooter")

Summarize your scRNA-seq data

A list of annotated Seurat objects can be summarized into a list of scoot objects using the scoot function. Compositional cell type distribution and aggregated transcriptomic profile (pseudobulk) are returned for each sample.

obj.list <- SplitObject(obj, split.by = "Sample")

scoot_object_list <- scoot(obj.list)

scoot_summary <- merge_scoot_objects(scoot_object_list)

scoot object content

The scoot object summarize the cell type annotation and contain the following slots:

• Seurat object metadata (dataframe): metadata
• Cell type composition for each layer of cell type prediction: composition. Including:
  • Cell counts
  • Frequency
  • CLR (Centred log ratio)-transformed counts (useful for downstream analyses such as PCA/Logratio analysis )
• Aggregated profile of predicted cell types: aggregated_profile. Including:
  • Aggregated expression per cell type.
  • Mean of UCell scores per cell type, if additional signatures are provided, for example from SignatuR.