The Waters You Delete Can Matter: A Safer GROMACS Cleanup Workflow

A molecular system can look ready for simulation and still contain structural details that complicate preparation: alternate atom locations, unwanted small molecules, crystal waters, or monatomic ions. Removing everything that looks extra may seem like the quickest fix—but some molecules, especially waters near an active site, may be important to the system you want to study.

A safer approach is to treat cleanup as a decision about what your simulation should represent. Before deleting anything, identify the molecules you intend to keep, then remove only what is unnecessary for that model. This is particularly important when preparing a protein–ligand system or working with a structure that contains functional waters.

Start with the structure you plan to simulate

In SAMSON, load a structure such as 1AKI using Home > Fetch, or open a structure saved on your computer. Then choose Home > Prepare to review the available cleanup options.

The Home > Prepare dialog in SAMSON

Removing alternate locations can help simplify the starting structure. For small molecules, make the choice based on the simulation you intend to run: use Remove ligands only when those ligands are not meant to be part of the model. If a ligand or cofactor should remain, do not remove it as part of a routine cleanup.

There is an important practical distinction here. A molecule may be present in the structure but still need a topology before it can be included in a GROMACS simulation. If you want to keep an arbitrary molecule, the documentation notes that you will need a topology file, such as an .itp. Tools such as the ATB Server or CGenFF can be used to generate files for molecules, which can then be provided during the GROMACS Wizard preparation step. If you do not need a molecule and cannot provide the required topology, removing it may be appropriate.

Don’t remove every water by default

Water removal is a choice that depends on the system. A broad removal step may be unsuitable when waters are tightly bound or take part in the function of an active site. If those waters matter to the model, preserve them and remove only waters outside the region you want to retain.

To do this, first select a structure, residue, or atom in the active site. In the Document view or Viewport, right-click the selection and choose Expand selection > Advanced. Set the node type to Water, then choose waters beyond a distance from the active-site selection. The auto-update option can help you inspect the selection before confirming it. Check the selected waters carefully, then use Erase selection to remove them.

Selecting water molecules beyond 5 Å from an active-site selection

The key safeguard is to verify the selection before erasing it. The distance defines which waters are selected relative to your chosen active-site atoms or structures; it is not a substitute for deciding which waters your scientific question requires. Adjust the selection to suit the system, and inspect it rather than assuming a single cutoff is right for every model.

A short pre-flight checklist

  • Alternate locations: remove them to simplify the structure.
  • Ligands and other small molecules: keep anything you intend to simulate; remove only molecules you do not need.
  • Water: decide whether a broad removal is appropriate. If active-site waters may be functional, select and remove only waters outside the region of interest.
  • Monatomic ions: remove them during this cleanup step as described in the workflow.
  • Hydrogens: adding them here is optional, since GROMACS can also add hydrogens.

After cleanup, check that the structure contains the molecules you intend to model. This makes the next stage—defining the simulation model, box, solvent, ions, and project files—a more deliberate transition rather than a last-minute attempt to recover something removed too early.

For the full walkthrough, including the preparation workflow, see the GROMACS Wizard pre-processing documentation.

Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.

Comments are closed.