PeptideTempVAE

This repostory is the implementation of the following paper:

Designing Novel Functional Peptides by Manipulating a Temperature in the Softmax Function Coupled with Variational Autoencoder

Reproduce the result by simply git clone and run python file

git clone https://github.com/shuan4638/PeptideTempVAE.git
python PeptideTempVAE.py

Framework Overview

<img src="https://github.com/shuan4638/PeptideTempVAE/blob/master/PeptideVAE.jpg">

Latent vectors encoded from generated peptides back to latent space

<img src="https://github.com/shuan4638/PeptideTempVAE/blob/master/Reencoded_peptides.png">

Authors

This code was written by Shuan Chen (PhD candidate of KAIST CBE) for efficient peptide design in 2019 for 2019 IEEE BigData Conference.