A Deep Learning Framework for Cell Type Prediction Fusing of Multi-angle Feature Sets from Single-cell Hi-C Data
The folder "Figure" contains the framework figure of scHiClassifier and the architecture figure of the fusion prediction model.
The folder "4DN" contains the data and code details for the 4DN dataset.
The folder "Collombet" contains the data and code details for the Collombet dataset.
The folder "Flyamer" contains the data and code details for the Flyamer dataset.
The folder "Lee" contains the data and code details for the Lee dataset.
The folder "Nagano" contains the data and code details for the Nagano dataset.
The folder "Ramani" contains the data and code details for the Ramani dataset.
It is recommended to deploy and run this project using PyCharm. We have also built an executable compute capsule for our code on the cloud-based platform Code Ocean, accessible at https://codeocean.com/capsule/2365852/tree. Due to circumstances on our end and the codeocean website, we are currently unable to adjust this instance to a public state, meaning it cannot be accessed by all accounts. Therefore, we suggest that when you visit "https://codeocean.com/capsule/2365852/tree", please log in using "202215269@mail.sdu.edu.cn" as the username and "captain@0827" as the password. After a successful login, you can run our deployed instance by clicking the "Reproducible Run" button at the top right corner of the page.
Mainly used libraries:
Python 3.9.15
torch 1.13.0
torchvision 0.14.0
sklearn
numpy
See "requirements.txt" for all detailed libraries.
Other developers can use the following command to install the dependencies contained in "requirements.txt":
pip install -r requirements.txt