MolDraw › Chemistry tools › PDB Cleaner
PDB Cleaner for Protein Preparation
In short: Load a PDB file or PDB ID, see its chains, waters, ions, ligands, altlocs and missing residues, then download a clean receptor and the ligand as separate PDB files. It is free and runs in your browser.
The first step before docking: strip what the docking program should not see and keep what it should. Default settings remove waters, single-atom ions and common crystallisation additives, keep one altloc per residue, and extract the largest ligand.
Structure summary
Cleaning options
Remove
Fix
Chains to keep
Ligands and other hetero groups
Extract a ligand to its own file
Metal ions in the active site (e.g. catalytic Zn²⁺) are usually part of the receptor. Untick “Single-atom ions” to keep them.
Cleaned receptor (.pdb)
Extracted ligand (.pdb)
What the cleaner does
A crystal structure from the PDB holds more than the protein: water, buffer ions, glycerol or sulfate from the crystallisation drop, alternate side-chain positions and sometimes several copies of the protein. Docking programs such as AutoDock Vina treat everything in the receptor file as rigid protein, so these extras must be removed or chosen deliberately.
| Group | How it is recognised | Default |
|---|---|---|
| Water | Residue name HOH, WAT, H2O, DOD, TIP3, SOL | Removed |
| Ions | Single-atom groups such as NA, K, CL, MG, CA, ZN, MN, FE, CU | Removed |
| Additives | SO4, PO4, GOL, EDO, PEG, ACT, DMS, MPD, CIT, FMT, TRS, MES … | Ticked for removal |
| Ligands | Any other HETATM group not in the polymer sequence | Kept; the largest is extracted |
| Modified residues | MSE, SEP, TPO, PTR, CSO … or any residue listed in SEQRES | Kept as protein |
Alternate locations
For each residue with altlocs, the tool keeps the label with the highest mean occupancy. If occupancies are equal it keeps the first label (A). Choosing per residue, not per atom, keeps each side chain in one consistent conformation. The altloc column is then blanked.
Missing residues
REMARK 465 lists residues that were in the sample but not seen in the electron density. If a file has no REMARK 465 (many edited files do not), the observed residues are aligned to the SEQRES sequence instead. Missing atoms inside residues (REMARK 470) are not checked.
How to prepare a protein for docking
- Load the PDB file, upload it, or import it by PDB ID.
- Read the summary. Note missing residues near the binding site and any metal that belongs to the active site.
- Keep one chain unless the pocket sits between chains. Remove waters, additives and ligands you do not need.
- Extract the co-crystallised ligand. Use it to set the Vina grid box and as a redocking control.
- Assign protonation states at your pH, add hydrogens, then convert to PDBQT.
Limitations
- It does not add hydrogens, charges, missing atoms or missing loops. Use PDBFixer, MODELLER or your docking suite for those.
- It reads PDB format only, not mmCIF. Very large entries without a PDB-format file cannot be imported by ID.
- Ion and additive detection uses residue-name lists. A rare additive is listed as a ligand; tick it to remove it.
- Cofactors (HEM, NAD, FAD …) are treated as ligands and kept by default, which is usually what you want.
- Atom serial numbers above 99,999 wrap around, as in standard PDB files.
Frequently asked questions
Should I remove water molecules before docking?
Usually yes. AutoDock Vina and most docking programs treat the receptor as rigid, so crystal waters block the pocket. Keep a water only if it bridges the ligand and protein in many structures of the same target, and then keep it on purpose.
What does keep highest-occupancy altloc mean?
Some atoms have two or more alternate positions (altloc A, B …) with occupancies that add up to 1. Docking programs need one position. For each residue the tool keeps the altloc with the highest mean occupancy (ties go to A) and blanks the altloc column.
How are missing residues found?
If the file has REMARK 465 records, those are used. Otherwise the observed residues are aligned to the SEQRES sequence. Without either, only gaps in residue numbering can be reported.
Does the cleaner add hydrogens or charges?
No. It only removes or keeps atoms and renumbers them. Add hydrogens and assign protonation states next, then convert the receptor and ligand to PDBQT.
Is my structure uploaded to a server?
No. Parsing and cleaning run in your browser. Only the optional PDB ID import downloads the file from RCSB.
Last updated: