scop: Single-Cell Omics analysis Pipeline

Introduction

The scop package provides a unified and extensible framework for single-cell omics data processing and downstream analysis in Seurat:

Integrated single-cell quality control methods, including doublet detection methods (scDblFinder, scds, Scrublet, DoubletDetection) and ambient RNA decontamination via decontX.
Pipelines embedded with multiple methods for normalization, feature reduction (PCA, ICA, NMF, MDS, GLMPCA, UMAP, TriMap, LargeVis, PaCMAP, PHATE, DM, FR), and cell population identification.
Pipelines embedded with multiple integration methods for scRNA-seq, including Uncorrected, Seurat, scVI, MNN, fastMNN, Harmony, Scanorama, BBKNN, CSS, Coralysis, LIGER, Conos, ComBat.
Multiple methods for automatic annotation of single-cell data (CellTypist, SingleR, Scmap, KNNPredict) and methods for projection between single-cell datasets (CSSMap, PCAMap, SeuratMap, SymphonyMap).
Multiple single-cell downstream analyses:
- Differential expression analysis: identification of differential features, expressed marker identification.
- Enrichment analysis: over-representation analysis, GSEA analysis, dynamic enrichment analysis.
- Cellular potency: CytoTRACE 2 for predicting cellular differentiation potential.
- RNA velocity: RNA velocity, PAGA, Palantir, CellRank, WOT.
- Trajectory inference: Slingshot, Monocle2, Monocle3, identification of dynamic features.
- Cell-Cell Communication: CellChat, CellphoneDB, NicheNet, and MultiNicheNet.
High-quality data visualization methods.
Fast deployment of single-cell data into SCExplorer, a shiny app that provides an interactive visualization interface.

scop: Single-Cell Omics analysis Pipeline

Credits

The scop package is developed based on the SCP package, with the following major improvements:

Compatibility: full support for Seurat v5.
Stability: a large number of known issues have been fixed, and all functions have passed devtools::check().
Usability: the Python environment setup workflow has been improved, allowing a new complete environment to be deployed within minutes; standardized console messages via thisutils::log_message for consistent, readable function outputs.
Performance: a new parallel framework has been developed based on thisutils::parallelize_fun, providing a consistent experience across Linux, macOS, and Windows.
Functionality: more analysis methods have been added, including CellRank, CellTypist, CytoTRACE 2, CellChat, GSVA, scMetabolism and scProportionTest.

Installation

R requirement

R >= 4.1.0

You can install the latest version of scop with pak from GitHub with:

if (!require("pak", quietly = TRUE)) {
  install.packages("pak")
}
pak::pak("mengxu98/scop")

Python environment

To run functions such as RunPAGA(), RunSCVELO(), scop requires conda to create a separate environment. The default environment name is "scop_env". You can specify the environment name for scop by setting options(scop_envname = "new_name").

Now, you can run PrepareEnv() to create the python environment for scop. If the conda binary is not found, it will automatically download and install miniconda.

scop::PrepareEnv()

To force scop to use a specific conda binary, it is recommended to set reticulate.conda_binary R option:

options(reticulate.conda_binary = "/path/to/conda")
scop::PrepareEnv()

If the download of miniconda or pip packages is slow, you can specify the miniconda repo and PyPI mirror according to your network region.

scop::PrepareEnv(
  miniconda_repo = "https://mirrors.bfsu.edu.cn/anaconda/miniconda",
  pip_options = "-i https://pypi.tuna.tsinghua.edu.cn/simple"
)

Pipeline

Data exploration

The analysis is based on a subsetted version of mouse pancreas data.

library(scop)
data(pancreas_sub)
print(pancreas_sub)
#> An object of class Seurat 
#> 47994 features across 1000 samples within 3 assays 
#> Active assay: RNA (15998 features, 0 variable feat

Scop

Install / Use

README