Sequence Modeling of Proteins using BiLSTM

A bidirectional long short-term memory RNN trained on the Protein Data Bank (PDB) dataset; the dataset was split into a 75,457 training set and 8,385 testing set. The dataset exceeds GitHub's 100MB limit, so only an example file is present. The current implementation includes functionality for sequence generation and long-distance dependency testing. Please refer to the materials folder for a more detailed write-up of the model.