Response to the comment on 'protein-protein binding affinity prediction from amino acid sequence'