Structures are pre-processed differently depending upon their input format (see the flowchart in Figure 1). Two input formats are supported by the H++ website, PDB and PQR. These files differ in that PQR files already have charges and radii assigned to each atom whereas PDB files do not.
If the input file is in PQR format, H++ makes minimal changes because it is assumed that the format is already suitable for electrostatic calculations. Changes made are as follows: atom names of all titratable amino acids are brought into accordance with the format adopted by the AMBER (29 ) package and consistency checks are performed. These checks ensure that the total charge of the system is an integer (within a ±0.05 unit charge tolerance per amino acid) and the atomic radii are between 0.5 and 3 Å. If any of the above checks fail, the sequence of residues is discontinuous or the atom names are different from the PDB standard and cannot be recognized, execution terminates.
For a structure submitted in the conventional PDB format, H++ deletes all HETATM records; that is, only those atoms that belong to amino acids or nucleotides are kept. This is the ‘clean-up’ step in Figure 1. Removal of explicit water molecules and mobile counterions is generally consistent with the implicit solvent framework in which solvation effects are accounted for in the mean-field manner. If necessary, removed ligands can be included in the calculations by submitting the complete structure in the PQR format, avoiding the ‘clean-up’ step. Sequence continuity is verified and all atom names are brought into accordance with the format adopted by the AMBER package. Deuterium atoms are replaced with equivalent hydrogens.