The nucleotide and protein features script is used to build the feature table data using the high confidence and low confidence genes. The two feature tables are used to train and test the Truguene protein and nucleotide classifier models.
There are two scripts which calculate feature tables given a nucleotide or protein fasta file: 'Ntd features.py' and 'Protein features.py'
These feature tables are given to two Jupyter notebooks. Pretrained weights are given in 'model_nucleotide.txt' and 'model_protein.txt'