SpaMosaic

SpaMosaic is a Python package for spatial multi-omics data integration using contrastive learning and graph neural networks. It supports integration of partially overlapping modalities and facilitates downstream analyses such as spatial domain identification and modality imputation.

🔧 Features

• Horizontal integration: integrates multiple slices within a single modality
• Vertical integration: integrates multiple modalities measured from the same slice
• Mosaic integration: integrates multiple slices with overallping modalities
• Imputation: imputes expression profiles of missing omics

🚀 Installation

Required Dependencies

SpaMosaic requires external installation of the following packages (not installed automatically):

• PyTorch (version ≥ 2.0)
• PyTorch Geometric (torch-scatter, torch-sparse, etc.)
• harmony-pytorch (version ≥ 0.1.7)

Once these dependencies are installed, simply run:

pip install spamosaic

⚠️ Note Both CPU and GPU versions of PyTorch and PyTorch Geometric are compatible with SpaMosaic.

📚 Documentation

📖 Full tutorials and API reference: 👉 https://spamosaic.readthedocs.io

📄 License

SpaMosaic is released under the MIT License. © 2025 Jinmiao Lab