NeuralEE

This is an applicable version for NeuralEE.

• The datasets loading and preprocessing module is modified from scVI, which is write for cpu computation.

• Define NeuralEE class and some auxiliary function, mainly for cuda computation, except like entropic affinity calculation which is quite faster computed on cpu.

• General elastic embedding algorithm on cuda is given based on matlab code from Max Vladymyrov.

• Add some demos of notebook helping to apply.

Usage

Installation

Installation of Pytorch

If you have an NVIDIA GPU, be sure to install a version of PyTorch that supports it. NeuralEE runs much faster with a discrete GPU.

Installation of NeuralEE

from GitHub

git clone git://github.com/HiBearME/NeuralEE.git
cd NeuralEE  
python setup.py install --user

Interactive command line

Reference from notebook files.

Experiments

CORTEX

Top left: elastic embeding with lambda=1.
Top right: elastic embeding with lambda=10.
Bottom left: NeuralEE with lambda=1.
Bottom right: NeuralEE with lambda=10.

HEMATO

Top left: elastic embeding with lambda=1.
Top right: elastic embeding with lambda=10.
Bottom left: NeuralEE with lambda=1.
Bottom right: NeuralEE with lambda=10.

PBMC

Top left: elastic embeding.
Top right: NeuralEE.
Bottom left: NeuralEE with 2 batches.
Bottom right: NeuralEE with 4 batches.

RETINA

Top left: NeuralEE with lambda=1.
Top right: NeuralEE with lambda=10.
Medium left: NeuralEE with 2 batches and lambda=1.
Medium right: NeuralEE with 2 batches and lambda=10.
Bottom left: NeuralEE with 4 batches and lambda=10.
Bottom right: NeuralEE with 4 batches and lambda=10.

BRAIN LARGE