**A Python/Dash implementation for comprehensive single-cell RNA sequencing analysis, maintaining architectural parity with the R/Shiny version.sequencing analysis, maintaining architectural parity with the R/Shiny version.
- Overview
- Key Features
- Why MASIH-Python?
- Quick Start
- Installation
- Documentation
- Analysis Modules
- Supported Data Formats
- Use Cases
- System Requirements
- Contributing
- Citation
- Roadmap
- Acknowledgments
- License
MASIH-Python provides a user-friendly web interface for analyzing single-cell RNA sequencing (scRNA-seq) data, with a focus on the multidimensional biological portrait of each cell.
Built with Dash and Scanpy, it offers a complete workflow from raw 10X Genomics data or processed AnnData objects to comprehensive biological insights.
- Accepts 10X Genomics outputs, H5AD files, and expression matrices
- Preserves existing analysis results and metadata
- Automatically detects and completes missing analysis steps
- Customizable quality control with interactive parameters
- High-Resolution Clustering — Leiden and Louvain algorithms with resolution optimization
- Dimensionality Reduction — PCA, t-SNE, UMAP for revealing data structure
- Marker Gene Profiling — Multiple differential expression methods with statistical validation
- Interactive Visualization — Plotly-powered plots with real-time exploration
- Functional State Characterization — CancerSEA-based scoring of 14 cancer-related pathways
- Trajectory Mapping — PAGA-based pseudotime analysis for developmental progression
- Cell Cycle Deconvolution — G1/S/G2M phase scoring integrated into downstream analyses
- Comparative Analysis — Cross-cluster and cross-pathway correlation studies
- Export publication-quality plots (PNG, PDF, SVG)
- Comprehensive data export (Excel, CSV, H5AD)
- Auto-generate dataset-specific methods text for manuscripts
- Batch export of all visualizations
- 🐍 Python Ecosystem: Leverage the power of Scanpy and the Python scientific stack
- 🌐 Web-Based: No local software installation required (with Docker)
- 🔬 Cancer-Focused: Specialized tools with CancerSEA pathway integration
- 💻 No Coding Required: Accessible to all researchers through intuitive interface
- 📊 Comprehensive Workflow: From raw data to publication-ready figures
- 🔄 Reproducible: Exports parameters and generates methods text
- 🧩 Modular Design: Easily extend with new analysis modules
- 🐳 Docker Support: One-command deployment for easy setup
# Option 1: Using Docker (Recommended)
git clone https://github.com/msherafatian/masih-python.git
cd masih-python
docker compose up --build
# Option 2: Using pip
git clone https://github.com/msherafatian/masih-python.git
cd masih-python
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -r requirements.txt
python app.py# If installed locally
python app.py
# If using Docker
docker compose upAccess the application at http://localhost:8050
Download example datasets from the Upload & QC tab, or use your own:
- 10X Genomics Cell Ranger output
- AnnData H5AD files
- Expression matrices (CSV/TSV)
- 📤 Data Upload: Multiple format support (10X, H5AD, CSV/TSV matrices)
- ✅ Quality Control: Interactive filtering with customizable thresholds
- 🎯 Clustering: Graph-based clustering with Leiden/Louvain algorithms
- 🧬 Marker Genes: Statistical testing with multiple methods (Wilcoxon, t-test, logreg)
- 🦠 CancerSEA Integration: 14 functional state pathways for cancer analysis
- 📊 Pathway Comparison: Correlation and comparative pathway analysis
- 🔀 Trajectory Analysis: PAGA-based pseudotime inference
- 🔄 Cell Cycle Scoring: G1/S/G2M phase identification and integration
- 📈 Interactive Plots: Real-time exploration with Plotly
- 🖼️ High-Quality Export: Publication-ready figures (PNG, PDF, SVG)
- 💾 Comprehensive Data Export: Excel, CSV, and H5AD formats
- 📝 Methods Generation: Automatic methods text for manuscripts
- 10X Genomics: Cell Ranger outputs (matrix directories, H5 files)
- AnnData Objects: H5AD files from Scanpy or other Python tools
- Expression Matrices: CSV/TSV format (genes × cells or cells × genes)
- Previous MASIH-Python Sessions: Reload processed data seamlessly
- 📤 Upload 10X data → 2. ✅ Quality control → 3. 🎯 Clustering → 4. 🦠 CancerSEA analysis → 5. 💾 Export results
- 🔄 Basic workflow → 2. 🔍 Cell type selection → 3. 🔀 Trajectory inference → 4. ⏱️ Pseudotime analysis → 5. 📊 Publication figures
- 📤 Load dataset → 2. 🎯 Clustering analysis → 3. 📊 Comparative pathways → 4. 📈 Statistical analysis → 5. 💾 Export results
MASIH-Python is designed for researchers studying:
- 🔬 Tumor Heterogeneity: Identify and characterize cancer cell subpopulations
- 💊 Treatment Response: Analyze single-cell responses to therapy
- 📈 Cancer Progression: Trace developmental trajectories and metastasis
- ⚡ Functional States: Characterize stemness, invasion, drug resistance
- 🧫 Microenvironment: Analyze tumor-immune interactions
- 🧬 Gene Expression Patterns: Discover marker genes and regulatory networks
- Python: Version 3.10 or higher
- Operating System: Windows 10+, macOS 10.14+, or Linux
- Memory: 4GB RAM minimum (8GB+ recommended for large datasets)
- Storage: 2GB free space for installation and dependencies
- Docker Desktop: Latest version
- Memory: 4GB RAM allocated to Docker (8GB+ recommended)
- Storage: 5GB free space for Docker images
We welcome contributions! Please read our Contributing Guidelines for details (coming soon).
- 🐛 Report bugs and request features
- 📖 Improve documentation
- 🧪 Add new analysis modules
- 🎨 Enhance user interface
- 🧬 Add new pathway databases
- 🐳 Improve Docker deployment
If you use MASIH-Python in your research, please cite the Zenodo DOI:
Concept DOI (latest MASIH-Python release):
https://doi.org/10.5281/zenodo.17824054
Version-specific DOI (e.g., for the archived release used in a paper):
vX.Y.Z → https://doi.org/10.5281/zenodo.17824053
@software{masih_python_2024,
author = {Sherafatian, Masih},
title = {MASIH-Python: Modular Analysis Suite for Interactive Heterogeneity},
year = 2024,
publisher = {Zenodo},
doi = {10.5281/zenodo.17824054},
url = {https://doi.org/10.5281/zenodo.17824054}
}
*[Full publication citation will be added upon publication]*- 📖 Documentation: Check our comprehensive guides (coming soon)
- 🐛 Issues: Report bugs on GitHub Issues
- ✉️ Email: masihshrftn@gmail.com
- 💬 Discussions: Join our GitHub Discussions
MASIH-Python is built on the shoulders of giants:
- Scanpy — Single-cell analysis in Python
- Dash — Web application framework by Plotly
- CancerSEA — Cancer functional state database
- CellRank — Trajectory inference
- Plotly — Interactive visualizations
- MASIH R/Shiny — Original implementation
Special thanks to all contributors and the single-cell community!
This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.
- MASIH (R/Shiny) — Original R implementation
- Scanpy — Single-cell analysis toolkit
MASIH-Python: Making single-cell cancer analysis accessible to all researchers.
