IJCAI2016

This repository is for the replication of our published paper Protein Secondary Structure Prediction Using Cascaded Convolutional and Recurrent Neural Networks on IJCAI2016. Our whole pipeline is listed as follows.

Download data

For cb513+profile_split1.npy.gz, cullpdb+profile_6133_filtered.npy.gz, please download from this website.
For CASP10 and CASP11, please download from this website.
Download data and put them in ./data folder.

Project Settings

1. Install the requirements (you can use pip or Anaconda):

   conda install pip h5py cython numpy scipy
   conda install -c conda-forge theano 
   conda install -c toli lasagne
   

2. You can do training/validation/test through this IPython notebook file Train_validation_test_release.ipynb.
   PS: The demo code use splited cullpdb+profile_6133_filtered for training/validation and then test on CB513 and CASP dataset. You can use whole cullpdb+profile_6133_filtered for training to obtain better performace.

Progress

• README for training
• README for project settings
• Dynamic training codes
• Dynamic evaluation codes
• Multi-GPU support

Acknowledgement

We thank Jian Zhou and Sheng Wang for CASP dataset generation.

Reference

If you find this code useful for your research, please cite

@inproceedings{li2016protein,
  title={Protein secondary structure prediction using cascaded convolutional and recurrent neural networks},
  author={Li, Zhen and Yu, Yizhou},
  booktitle={Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence},
  pages={2560--2567},
  year={2016},
  organization={AAAI Press}
}