Course project for CS 273B and CS 229. The goal is to train a deep learning model that takes one of eight kinases and a potential inhibitor, and predicts whether or not the kinase will be inhibited
- Download the dataset from https://www.kaggle.com/xiaotawkaggle/inhibitors/home
- Unzip and rename to "data", place in the repo's base directory
- Make sure your conda version is 4.3.34
- Create a conda virtual environment https://conda.io/docs/user-guide/tasks/manage-environments.html
- conda create -n myenv python=3.5.3
- activate
- sudo apt-get install -y libxrender-dev
- pip install -r requirements.txt
- conda install cython
- conda install -c conda-forge mdtraj
- Install rdkit https://www.rdkit.org/docs/Install.html sudo apt-get install python-rdkit librdkit1 rdkit-data
- Install https://github.com/kundajelab/simdna
- gpu=0 bash scripts/install_deepchem_conda.sh (gpu=1 for GPU support)
- Move deepchem folder to base directory