MAPS is a modality-agnostic spatial omics computing platform that unifies physical and molecular coordinates systems across arbitrary modalities, enabling ultrafast cross-modal alignment, unpaired diagonal integration across orthogonal modalities, paired 3D multimodal reconstruction, and 3D interactive exploration.
git clone https://github.com/dbjzs/MAPS.git
cd MAPS
MAPS is available for Python 3.10. We recommend to train MAPS models on a device with GPU support.
- Using the conda install environment
conda create -n MAPS -c conda-forge python==3.10.13 libopenblas=0.3.25 -y
conda activate MAPS
- Then using pip local install MAPS.
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install .
MAPS is available for Python 3.10. We recommend to train MAPS models on a device with GPU support.
- Using the conda install environment
conda create -n MAPS -c conda-forge python==3.10.13 libopenblas=0.3.25 -y
conda activate MAPS
- Then using pip install MAPS.
pip install mapspatial
- All h5ad files have been uploaded to the zenodo repository
- Jupyter tutorials and API documentation are available at Tutorial.
- The expected data input format of MAPS is AnnData object.
- Please use issues to submit bug reports.
- Jupyter Tutorial
Please use the core name as follows:
pip install ipykernel python -m ipykernel install --user --name=MAPS --display-name="Python (MAPS)"Python [conda env:MAPS]
- 3D interactive exploration website is available at MAPS-Explorer.
- 3D analysis tutorial 3D analysis.
- If you find MAPS useful for your research, please consider citing the MAPS manuscript bioRxiv.
@article{,
title = {Unifying physical and molecular coordinates systems across modalities in spatial biology},
author = {Dai, Yan, Wang*, Liang,....Zuo*, Qian*, Yuan* et al.},
year = {2026},
journal = {bioRxiv : Unifying physical and molecular coordinates systems across modalities in spatial biology},
eprint = {https://www.biorxiv.org/content/10.64898/2026.09.01.748536v1},
doi = {https://doi.org/10.64898/2026.09.01.748536},
}