Predicting Protein Secondary Structure using RNN & CNN

Model tested on the CB513 dataset.

Note: This code is optimised to run on Tensorflow version 1.3.0.

Motivation

To build a RNN-based neural network model capable of predicting the secondary structure of proteins based on their amino-acid chains.

Model architecture

TODO

Results

TODO

This repository represents the author's submission for COMP3212: Computational Biology course at the University of Southampton