steps to install

[Rohit-Satyam]

The following installation steps worked for me:

mamba create -n temp3 notebook pytorch=1.4.0 -c pytorch -c conda-forge
pip install biopython
pip3 install requests
mamba install pyrsistent
pip install sklearn

[Rohit-Satyam]

Caution:
Few changes needs to be made in the script because the numpy arrays refuses to get saved until dtype=object is specified in parse_entries.py.

                #np.save(dir_input + 'proteins_' + str(10*q_count+1) + '_' + str(10*(q_count+1)), proteins)
                #np.save(dir_input + 'adjacencies_' + str(10*q_count+1) + '_' + str(10*(q_count+1)), adjacencies)
                #np.save(dir_input + 'names_' + str(10*q_count+1) + '_' + str(10*(q_count+1)), pnames)
                #np.save(dir_input + 'seqs_' + str(10*q_count+1) + '_' + str(10*(q_count+1)), pseqs)
                np.save(os.path.join(dir_input, f'proteins_{10*q_count+1}_{10*(q_count+1)}'), np.array(proteins, dtype=object))
                np.save(os.path.join(dir_input, f'adjacencies_{10*q_count+1}_{10*(q_count+1)}'), np.array(adjacencies, dtype=object))
                np.save(os.path.join(dir_input, f'names_{10*q_count+1}_{10*(q_count+1)}'), np.array(pnames, dtype=object))
                np.save(os.path.join(dir_input, f'seqs_{10*q_count+1}_{10*(q_count+1)}'), np.array(pseqs, dtype=object))

[Rohit-Satyam]

In addition, change the Kfold line to the following in k-fold-CV.py:

kfold=KFold(n_splits=num_kfolds, shuffle=True, random_state=1)