A docking run can fail—or produce results that are difficult to interpret—before the ligand ever reaches the binding site. Protein files may contain alternate atom positions, unwanted small molecules, missing atoms, or hydrogens that have not been added. Cleaning everything automatically is not always the right answer either: waters, ions, and cofactors can matter to the biological question.
A useful preparation workflow starts with a simple question: What does this particular downstream task require? Then choose the lightest set of cleanup and repair steps that meets that need, and validate the resulting structure before moving on.
Start with the structure you have
First decide whether the model appears to be a routine cleanup or needs structural repair. If it is already complete and looks reasonable, a quick preparation may be enough. If residues or atoms are missing, or you need hydrogens at a specified pH, use a deeper repair workflow rather than treating the job as simple file cleanup.
In SAMSON, Home > Prepare provides a one-step starting point. It can resolve alternate locations, delete selected ligands and other small molecules, strip waters, clear monatomic ions, and add hydrogens. Consider each operation in the context of your system: for example, a water molecule or cofactor that is unnecessary for one workflow may be relevant to another.
Resolve alternate locations before they become a hidden complication
Some structures list alternative positions for atoms. Keeping several alternatives can leave ambiguity in the model, while choosing one gives downstream tools a single set of coordinates to work with. SAMSON’s preparation step keeps the highest-occupancy atoms. When occupancies are equal or absent, it prefers alternate location A when available, then B, and so on.
If you want to inspect the choices first, open Home > Validate, select the Alt. locations tab, and click Find alternate locations. Review the list, then use Remove alt. locations to apply the selection. This is a useful point to pause and check what will be retained rather than assuming every structure has the same pattern.
Clean selectively, then check the geometry
Ligands, waters, and ions can be selected separately from Select > Biology. Delete only the components that are not needed for the planned calculation. The same restraint applies to hydrogens: Edit > Add hydrogens adds standard hydrogens at pH 7 for standard amino and nucleic acid residues, and uses valences for other structures. You can select a substructure first to limit where hydrogens are added.
Adding hydrogens does not automatically mean the whole system must be minimized. The appropriate follow-up depends on the workflow. For example, the documentation notes that AutoDock Vina needs only polar hydrogens to determine hydrogen bonds. Avoid adding extra steps by default; follow the requirements of the method you plan to use.
After cleanup, use Home > Validate to inspect bond lengths, non-standard residue names, and clashes. A detected clash is a prompt to investigate, not necessarily a reason to alter the structure blindly. For side-chain clashes, SAMSON’s Rotamers editor can change side-chain conformations using a backbone-dependent rotamer library.
When cleanup is not enough
If residues or heavy atoms are missing, use the PDBFixer extension for deeper repairs. It can add missing residues based on SEQRES records, build missing heavy atoms and side chains, convert non-standard residues to standard equivalents, and add hydrogens for a specified pH. It can also resolve alternate locations, remove water or heterogens, and build explicit-solvent water boxes or lipid membranes for suitable systems.
For repeated work, Batch Protein Prepare can process a folder of structures while preserving its subfolder structure. It accepts PDB, PDBx/mmCIF, MMTF, and MOL2 inputs, and can also download and prepare structures from PDB identifiers.
The practical takeaway: distinguish cleanup from repair, keep biologically relevant components, and validate the model after making changes. For the full workflow and tool details, read the protein preparation and validation tutorial.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.