1. Chemical & Macromolecular File Formats

Molecular docking requires converting standard structural formats into specialized representation formats that define atom types, partial charges, and rotatable dihedral bonds.

2. Receptor Preparation Protocol: Resolving Experimental Artifacts

Raw PDB crystal structures cannot be docked directly due to experimental artifacts, unassigned hydrogens, and crystallographic additives. Follow this systematic preparation checklist:

• Crystallographic Water Molecules: Remove bulk crystallization solvent ($H_2O$ / HOH). Crucial exception: Conserved structural waters bridging catalytic residues and the ligand (e.g., $W_1$ in HIV-1 protease or Wat418 in viral polymerases) should be retained and docked as a receptor-water adduct.
• Heteroatoms & Co-Crystallized Additives: Remove cryoprotectants (glycerol, PEG), buffer salts (sulfates, phosphates), and native co-crystallized inhibitors (save the native ligand coordinates for grid box centering and redocking validation).
• Alternate Conformations (AltLoc): PDB files frequently contain multiple occupancies (e.g., atom records with AltLoc 'A' and 'B'). Select the higher occupancy conformation or conformer 'A'.
• Protonation States at Physiological pH 7.4: Amino acid sidechains titrate based on local electrostatic microenvironments. Use PROPKA 3.4 to predict $pK_a$ shifts: histidine tautomers ($\text{HID}$ with $\delta$-protonation, $\text{HIE}$ with $\epsilon$-protonation, or $\text{HIP}$ diprotonated positively charged state) must be assigned accurately.
• Partial Charges: X-ray diffraction does not resolve electrons on individual nuclei. Assign Kollman united-atom partial charges for the receptor.

# 1. Strip non-protein heteroatoms and crystallization waters
grep -v "^HETATM" target_raw.pdb | grep -v "HOH" > target_clean.pdb

# 2. Add polar hydrogens and compute protonation at pH 7.4
obabel -ipdb target_clean.pdb -opdb -O target_h.pdb -p 7.4

# 3. Convert to PDBQT with Kollman charges and AutoDock atom typing
prepare_receptor4.py -r target_h.pdb -o target_receptor.pdbqt -A checkhydrogens -U nphs_lps_waters

3. Small-Molecule Ligand Preparation: Conformation & Torsion Trees

Ligands are typically stored as 1D SMILES strings or 2D SDF connection tables. Converting them into 3D docking-ready structures requires energy minimization and defining rotational degrees of freedom:

# Convert 2D SMILES into 3D coordinates with MMFF94 force field minimization
obabel -:"CC(=O)Oc1ccccc1C(=O)O" -osdf -O aspirin_3d.sdf --gen3d --minimize --ff MMFF94 --steps 500

# Convert to PDBQT, calculate Gasteiger partial charges, and assign rotatable bonds
prepare_ligand4.py -l aspirin_3d.sdf -o aspirin.pdbqt -A hydrogens_charges

Inside the generated .pdbqt ligand file, the tags ROOT, BRANCH, and ENDBRANCH specify the torsion tree. Rigid aromatic bonds and double bonds are locked, while $sp^3-sp^3$ single bonds are allowed full 360° rotational freedom during search space exploration.