SuperCT

A PyTorch Implementation of SuperCT for Single-Cell Clustering

Ⓒ Copyright Tianyi Liu 2019, All Rights Reserved.

Latest Update: 15:50 Aug 20th, 2019 CST

💥If you use or partially use this repo, you shall formally - in report or presentation - acknowledge this repo.

Introduction

SuperCT was initially published by Xie et al., in Nucleic Acids Research Vol. 47, 2019. This repo presents a modified version of SuperCT for supervised single-cell clustering. We have tested B, M, and T cells clustering with the implemented model on an industrial dataset and achieved 99.28% accuracy during testing.

Structure of this repo

source contains all the source files

img contains the images used in this README.md

Usage

Input: Binary single-cells expression matrices

trainvalnet.py for training and validation

testnet.py for testing after training

cfg.py contains all hyperparameters, Make sure to modify this before you use this repo

Package Requirements

matplotlib

pickle

PyTorch

scikit-learn

Numpy

Description

The model was modified as shown above and implemented in PyTorch v1.0.

With ~4k8 cells of three categories, this repo achieved 99.28% accuracy and 0.9965 mAP when testing.