All code is in the root directory. Due to file size limitations of Github, data files are not included but they can be found at Drugbank.ca. File paths should be adjusted to the user's own file layout if code is run.
Notebooks are listed in the order they should be used in. THE 'OTHER FILES' SHOULD BE RUN BEFORE THESE ARE USED
- feature_extraction: Takes data and extracts protein and drug features.
- dataset_creation: Creates training and testing dataset from the outputs of the feature_extraction notebook
- lasso_training: Determines optimal lasso model and trains a lasso model for proteins and drugs. Uses the models to select features and saves the reduced datasets for use in machine learning model training.
- machine_learning_training: Trains various machine learning models (CNN, DNN, Random Forest) and shows metrics.
- covid_prediction: Uses the trained machine learning models to make predictions for SARS-CoV2 viral proteins.
- modify_fasta: Replaces the '|' in the fasta files with spaces for easier processing later. Should be run before the notebooks are used.
- struc_to_smiles: Extracts SMILES for approved drugs from DrugBank data and writes to a separate
.smifile. Used to calculate drug descriptors (using ochem.eu).