Sovereign 8s.is mirror of standardgalactic/DePass
  • Python 94.6%
  • Shell 3.1%
  • R 2.3%
Find a file
Repository files (latest commit first)
Filename Latest commit message Latest commit date
yuanxiangjiang b45dfa3bd8 “readme”
2026-09-08 17:24:05 +08:00
data “readme” 2026-09-08 17:24:05 +08:00
DePass up 2026-04-24 13:23:17 +08:00
docs “readme” 2026-09-08 17:24:05 +08:00
scripts Convert scMOG to normal directory 2026-04-24 23:17:02 +08:00
.readthedocs.yaml up 2026-04-24 13:23:17 +08:00
DePass-0.0.25-py3-none-any.whl updata 2026-04-24 23:12:05 +08:00
LICENSE updata DePass 2025-07-17 16:12:50 +08:00
readme.md “readme” 2026-09-08 17:24:05 +08:00
requirements.txt updata 2025-09-06 20:20:25 +08:00

DePass

python >3.8.20

DePass is a dual-enhanced graph learning framework designed for integrated analysis of both single-cell and spatial paired multi-omics data. It flexibly supports diverse modality combinations.

alt text


Tutorials

For detailed tutorials, you can refer to the following resources:


Downstream tasks and evaluations

We provide complete scripts for all downstream tasks and evaluations in this repository, structured as follows:

  • Spatial niche inference
  • Cell-cell communication analysis
  • Trajectory inference
  • Omics data prediction
  • Enhancement evaluation

All scripts are available in the scripts directory: Downstream scripts.


DePass Installation

It is preferred to create a new environment for DePass.

conda create -n DePass python==3.8.20
conda activate DePass

Install DePass from GitHub

git clone https://github.com/zhanglabNKU/DePass.git
cd DePass
pip install DePass-0.0.25-py3-none-any.whl

Additional Dependencies

Because DePass leverages mclust for clustering, installing R, the rpy2 Python interface, and the mclust R package is recommended.

conda install -c conda-forge r-base rpy2
conda install conda-forge::r-mclust

Install PyTorch and PyTorch Geometric

pip install torch==2.4.1
pip install torch-geometric==2.3.1
pip install torch_scatter torch_sparse torch_cluster torch_spline_conv -f https://data.pyg.org/whl/torch-2.4.0+cu121.html

The typical install time on a normal desktop computer should be less than 5 minutes.


Requirements

anndata==0.11.4
hnswlib==0.8.0
matplotlib==3.7.1
numpy==1.23.5
pandas==1.5.3
scanpy==1.11.3
scikit_learn==1.3.0
scipy==1.16.0
seaborn==0.13.2
setuptools==75.3.2
tqdm==4.65.0
torch==2.4.1
torch_geometric==2.3.1
torch_scatter==2.1.2+pt24cu121
torch-sparse==0.6.18+pt24cu121
torch-cluster==1.6.3+pt24cu121
torch-spline-conv==1.2.2+pt24cu121
R==4.3.1
rpy2==3.5.11
jupyter
mclust

Data

The data used in this study are available at Data.