SCP - Single Cell Pipeline (Seurat V5 Compatible)

An adaptation of the SCP package with full Seurat V5 support and modern Python dependency management.

Features

🎯 Seurat V5 Compatibility

• Full V5 Support: Native compatibility with Seurat V5's layer-based data structure
• Version Agnostic: Seamlessly handles both Seurat V4 and V5 objects
• Automatic Detection: Smart detection and conversion of legacy objects
• Zero Migration Cost: Existing pipelines work without modification

⚡ Fast Python Dependency Management

• UV-Only: Ultra-fast Python package manager (10-100x faster than pip/conda)
• Modular Installation: Install only the features you need
• Cross-Platform: Works on macOS, Linux, and Windows
• Simple Setup: One command to install all dependencies

🔬 Comprehensive Single-Cell Analysis

• RNA velocity analysis (scVelo, CellRank)
• Trajectory inference (Palantir, WOT)
• Advanced dimensionality reduction (UMAP, PHATE, PaCMAP, TriMAP)
• Batch correction (Scanorama, BBKNN, Harmony)
• Doublet detection (Scrublet, DoubletDetection)
• Deep learning methods (scvi-tools)
• And much more...

Installation

Install from GitHub

# Install devtools if needed
install.packages("devtools")

# Install SCP
devtools::install_github("jinworks/SCP-SeuratV5")

Python Environment Setup

SCP uses Python for advanced analysis features, managed exclusively via UV (ultra-fast package manager).

library(SCP)

# Install core dependencies only (~5 seconds)
PrepareEnv()

# Install all features (~60 seconds)
PrepareEnv(extras = "all")

# Install specific features
PrepareEnv(extras = c("velocity", "trajectory", "spatial"))

# Add more features to existing environment later
uv_install_extras("deeplearning")

Performance: UV is 10-100x faster than pip/conda!

For detailed Python setup instructions, see PYTHON_DEPENDENCIES.md.

Quick Start

Basic Workflow

library(SCP)

# Load your Seurat object (automatic V5 conversion if needed)
srt <- readRDS("your_seurat_object.rds")

# Check if it's Seurat V5
IsSeurat5(srt)

# Convert to V5 if needed
srt <- EnsureSeurat5(srt)

# Run standard analysis
srt <- RunStandardWorkflow(srt)

# Visualize
CellDimPlot(srt, group.by = "cell_type")
FeatureDimPlot(srt, features = c("CD3D", "CD8A", "CD4"))

Working with Seurat V5

# Version-agnostic data access
counts <- get_seurat_data(srt, layer = "counts")
data <- get_seurat_data(srt, layer = "data")

# Set data back
srt <- set_seurat_data(srt, data = normalized_data, layer = "data")

# Feature metadata
feature_meta <- get_feature_metadata(srt)
srt <- set_feature_metadata(srt, metadata = new_meta)

# Variable features
var_genes <- get_var_features(srt)

Python-Based Analysis

# RNA Velocity with scVelo
srt <- RunScvelo(srt,
  group.by = "cell_type",
  linear_reduction = "pca",
  nonlinear_reduction = "umap"
)

# Trajectory with Palantir
srt <- RunPalantir(srt,
  start_cell = "CELL_001",
  num_waypoints = 500
)

# Cell fate with CellRank
srt <- RunCellRank(srt,
  group.by = "cell_type",
  linear_reduction = "pca"
)

Requirements

R Dependencies

• R >= 4.1.0
• Seurat >= 5.0.0
• SeuratObject >= 5.0.0
• Plus standard Bioconductor packages (automatically installed)

Python Dependencies (Optional)

• Python 3.10, 3.11, or 3.12 (3.10 recommended)
• Managed via UV (ultra-fast package manager)
• Quick start:

# One-time install
PrepareEnv(extras = c("spatial", "velocity"))

# In new sessions, activate environment
library(SCP)
use_uv_env()

See PYTHON_DEPENDENCIES.md for the complete list.

Key Functions

Version Management

• IsSeurat5(srt) - Check if object is Seurat V5
• EnsureSeurat5(srt) - Convert to V5 if needed
• GetSeuratVersion(srt) - Get detailed version info

Data Access (Version-Agnostic)

• get_seurat_data(srt, layer) - Get data from any layer
• set_seurat_data(srt, data, layer) - Set data to a layer
• get_feature_metadata(srt) - Get feature metadata
• set_feature_metadata(srt, metadata) - Set feature metadata
• get_var_features(srt) - Get variable features

Python Environment

• PrepareEnv(extras) - Set up UV Python environment with optional feature groups
• use_uv_env() - Configure R to use UV environment
• check_uv() - Check if UV is installed
• uv_env_exists() - Check if UV environment exists
• uv_install_extras(extras) - Add optional feature groups to existing environment
• uv_install_packages(packages) - Install individual Python packages
• RemoveEnv() - Remove UV Python environment
• check_Python(packages) - Verify Python packages are available

Analysis Functions

See the original SCP documentation for the full list of analysis functions.

Examples/Tests

• Minimal scRNA smoke test: tests/testthat/test_seurat_v5_basics.R
• Optional Chromium test: set SCP_TEST_10X_H5 to a 10x HDF5 path
• Optional Visium HD test: set SCP_TEST_VISIUM_BASE to a binned output dir (e.g., .../square_008um) and run tests/interactive/test_spatial_visium_hd.R

Documentation

• Python Dependencies: PYTHON_DEPENDENCIES.md
• UV Setup Guide: UV_SETUP.md
• Original SCP Docs: SCP Repository

Testing Your Installation

Test Python Dependencies

library(SCP)

# Activate UV environment
use_uv_env()

# Check specific packages
check_Python(c("numpy", "pandas", "scanpy", "anndata"))

# Check UV installation
check_uv()
uv_env_exists()

Quick Test

library(SCP)

# Test basic functionality
IsSeurat5(your_object)

# Test Python environment
use_uv_env()
py_config <- reticulate::py_config()
print(py_config)

Troubleshooting

Python Environment Issues

# Remove and recreate environment
RemoveEnv()
PrepareEnv(force = TRUE)

# Or manually
uv_remove_env()
PrepareEnv(extras = "all")

# Verify installation
use_uv_env()  # Activate environment
check_Python(c("numpy", "scanpy", "anndata"))  # Check packages

Seurat V5 Conversion Issues

# Get detailed version info
GetSeuratVersion(srt)

# Force update to V5
srt <- UpdateSeuratObject(srt)
srt <- EnsureSeurat5(srt)

Missing Optional Dependencies

# Install specific feature groups to existing environment
uv_install_extras(c("velocity", "trajectory"))

# Or reinstall with new extras
PrepareEnv(extras = c("velocity", "trajectory", "spatial"), update = TRUE)

Performance Tips

1. Modular Installation: Install only specific extras you need instead of "all" for faster setup
2. Seurat V5 Native: Use V5 objects for better memory efficiency
3. Apple Silicon: Use extras = "apple_silicon" for Metal acceleration in deep learning
4. UV Speed: Python environment setup is 10-100x faster than traditional pip/conda

Platform Support

Platform	Python (UV)	Status
macOS (Intel)	✅	Full support
macOS (Apple Silicon)	✅	Full support + Metal acceleration
Linux	✅	Full support
Windows	✅	Full support

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

Citation

If you use SCP in your research, please cite the original package:

Zhang, H. (2024). SCP: Single Cell Pipeline.
GitHub: https://github.com/zhanghao-njmu/SCP

License

GPL-3.0-or-later

Acknowledgments

• Original SCP package by Hao Zhang
• Seurat team for Seurat V5
• UV team at Astral for the amazing Python package manager

Related Projects

• Seurat - Single-cell RNA-seq analysis
• scanpy - Python-based single-cell analysis
• scVelo - RNA velocity analysis
• UV - Fast Python package manager

Support

For issues and questions:

• Bug Reports: GitHub Issues
• Original SCP: SCP Issues
• Python Dependencies: See PYTHON_DEPENDENCIES.md

---

Version: 0.7.0 Last Updated: 2025-10-14