Celline - Single Cell RNA-seq Analysis Pipeline

Celline is a comprehensive, interactive pipeline for single-cell RNA sequencing (scRNA-seq) analysis, designed to streamline the workflow from raw data to biological insights. It provides both command-line and web-based interfaces for flexible analysis workflows.

📖 Detailed Documentation: Celline Docs

Features

• 🔄 Automated Data Processing: From raw FASTQ files to expression matrices
• ✅ Quality Control: Built-in QC metrics and filtering with Scrublet doublet detection
• 📊 Dimensionality Reduction: PCA, t-SNE, and UMAP implementations
• 🔍 Clustering Analysis: Multiple clustering algorithms
• 🧬 Cell Type Prediction: Automated cell type annotation using scPred
• ⚖️ Batch Effect Correction: Multiple methods for data integration (Seurat, scVI)
• 🌐 Interactive Visualization: Web-based interface for data exploration
• 🔧 Flexible Execution: Support for local multithreading and PBS cluster execution
• 📁 Database Integration: Built-in support for SRA, GEO, and CNCB data repositories
• 🔬 R Integration: Seamless R/Seurat integration for advanced analysis

System Requirements

Required Dependencies

• Python: ≥3.10
• R: ≥4.0 with Seurat and other required packages
• Cell Ranger: For 10x Genomics data processing
• SRA Toolkit: For downloading SRA data (fastq-dump)

Python Dependencies

All Python dependencies are automatically installed via pip. Key packages include:

• scanpy - Single-cell analysis
• pandas, polars - Data manipulation
• fastapi, uvicorn - Web API
• rich - Enhanced CLI interface
• pysradb - SRA database access

Installation

Option 1: Install from PyPI

pip install celline

Option 2: Install from Source

git clone https://github.com/your-repo/Celline.git
cd Celline
pip install -e .

Option 3: Development Installation

git clone https://github.com/your-repo/Celline.git
cd Celline
pip install -e ".[dev]"

Quick Start

1. Initialize Your Project

Start by initializing a new project. This will validate system dependencies and create configuration files:

celline init

This command will:

• Check for required system dependencies (R, Cell Ranger, SRA Toolkit)
• Set up R environment configuration
• Create project configuration files
• Prompt for project name and settings

2. Configure Execution Settings (Optional)

Configure execution parameters for your system:

# Interactive configuration
celline config

# Or set specific options
celline config --system multithreading --nthread 8
celline config --system PBS --pbs-server your-cluster-name

3. Explore Available Functions

List all available analysis functions:

celline list

Get detailed help for specific functions:

celline help download
celline help preprocess

4. Basic Analysis Workflow

Download Public Data

# Download from SRA/GEO
celline run download --accession GSE123456
celline run download --accession SRR123456

# Download from CNCB
celline run download --accession CRA123456

Data Preprocessing

# Quality control and preprocessing
celline run preprocess --input raw_data/ --output processed/

# Gene expression counting (10x data)
celline run count --input cellranger_output/ --output counts/

Create Seurat Objects

# Create Seurat object for downstream analysis
celline run create_seurat --input counts/ --output seurat_object.rds

Advanced Analysis

# Dimensionality reduction
celline run reduce --input seurat_object.rds --methods pca,umap,tsne

# Cell type prediction
celline run predict_celltype --input seurat_object.rds --reference ref_data/

# Batch effect correction
celline run integrate --input multiple_samples/ --method seurat

5. Interactive Web Interface

Launch the interactive web interface for visual analysis:

celline interactive

This will:

• Start the FastAPI backend server
• Launch the Vue.js frontend
• Open your web browser automatically
• Provide interactive data exploration tools

6. API Server Only (for Development)

Start only the API server for testing:

celline api

Available Functions

Project Structure

When you initialize a project, Celline creates the following structure:

your_project/
├── setting.toml          # Project configuration
├── data/                 # Raw and processed data
├── results/              # Analysis results
├── scripts/              # Generated analysis scripts
└── logs/                 # Execution logs

Configuration

Celline uses setting.toml files for configuration:

[project]
name = "my_project"
version = "0.01"

[execution]
system = "multithreading"  # or "PBS"
nthread = 8
pbs_server = "your-cluster"  # for PBS system

[R]
r_path = "/usr/local/bin/R"

[fetch]
wait_time = 4  # seconds between API calls

Advanced Usage

Running on HPC Clusters

For PBS/Torque clusters:

celline config --system PBS --pbs-server your-cluster-name
celline run preprocess  # Will submit PBS jobs automatically

Custom Analysis Scripts

Celline generates executable scripts in the scripts/ directory that can be run independently or modified for custom workflows.

R Integration

Access Seurat objects and run custom R analysis:

# R scripts are available in template/hook/R/
# Custom R functions can be added to the pipeline

Troubleshooting

Common Issues

1. Missing Dependencies: Run celline init to validate all dependencies
2. R Package Issues: Ensure Seurat and required R packages are installed
3. Memory Issues: Adjust thread count with celline config --nthread <number>
4. Web Interface Not Loading: Check that ports 8000 and 3000 are available

Getting Help

# General help
celline help

# Function-specific help
celline help <function_name>

# System information
celline info

# List all functions
celline list

Contributing

We welcome contributions! Please see our contributing guidelines for more information.

Citation

If you use Celline in your research, please cite:

[Citation information to be added]

License

This project is licensed under the MIT License - see the LICENSE file for details.

Support

• 📖 Documentation
• 🐛 Issue Tracker
• 💬 Discussions