Deep soft K-means clustering with self-training for single cell RNA sequence data.

Architecture

Requirement

Python 3.6

Tensorflow 1.14

Keras 2.2

Data availability

The real data sets we used can be download in data.

Quick start

We use the dataset “Bach” and ZINB distribution modelling to give an example. You just run the following code in your command lines:

python zidpkm.py --dataname "Bach"

Then you will get the cluster result of “Bach” dataset using scziDesk method in ten random seed. The median values of Accuracy, ARI and NMI are 0.9046, 0.8738 and 0.8343, respectively.

Reference

Our paper is published in NAR Genomics and Bioinformatics. The details can be seen in article. Please consider citing it.

Contributing

Author email: clandzyy@pku.edu.cn