Each protein chain is represented using six lines:
1: Protein ID
2: Amino acid sequence for the test dataset or selected residue for the training dataset
3: The DNA-binding annotations where 
    0 indicates non-DNA-binding
    1 indicates for DNA-binding annotations
    2 (only for the test dataset) shows residues that were not used since the binding information was unavailable for these residues
4: Space-separated RAA values for each residue
5: Space-separated putative RSA values for each residue (predicted with ASAquick)
6: Space-separated ECO value for each residue (computed using profiles generated with HHblits)