https://www.cell.com/cell-systems/fulltext/S2405-4712(22)00395-7
Fast-Higashi is an interpretable model that takes single-cell Hi-C (scHi-C) contact maps as input and jointly infers cell embeddings as well as meta-interactions.
We now have Fast-Higashi on conda!
Do
conda install -c ruochiz fasthigashi
or
mamba install -c ruochiz fasthigashi
After that install the latest pytorch with corresponding CUDA support. Check https://pytorch.org for details. Note that fasthigashi won't check if you have pytorch installed. So, the user would have to install the correct pytorch version individually.
git clone https://github.com/ma-compbio/Fast-Higashi/
cd Fast-Higashi
python setup.py install
It is recommended to have pytorch installed (with CUDA support when applicable) before installing higashi.
The input format would be exactly the same as the Higashi software. Detailed documentation will be updated here at the Higashi wiki
Cite our paper by
@article {Zhang2022fast,
author = {Zhang, Ruochi and Zhou, Tianming and Ma, Jian},
title = {Ultrafast and interpretable single-cell 3D genome analysis with Fast-Higashi},
year = {2022},
doi = {10.1016/j.cels.2022.09.004},
journal={Cell systems},
volume={13},
number={10},
pages={798--807},
year={2022},
publisher={Elsevier}
}
Please contact zhangruo@broadinstitute.org or raise an issue in the github repo with any questions about installation or usage.