[BUG] genion fails with crystallographic water in a protein-ligand system

[cespos]

Describe the bug
In my system, I have a protein with two crystallographic waters bound to a ligand.
BSS.Solvent.solvate() fails with the following error:

Looking into the genions.err, I found the error message:

And this is because in the input system, I have in order:

• protein
• crystallographic waters
• ligand

Therefore, when solvent waters are added, genions sees water molecules that are not continuous but with a ligand in between.
How to best deal with this?

To Reproduce

import BioSimSpace as BSS

protein_coord = BSS.IO.readMolecules('protein.pdb')[0]
ligand_coord = BSS.IO.readMolecules('ligand.sdf')[0]

protein = BSS.Parameters.parameterise(protein_coord , 'ff14SB', water_model='tip3p').getMolecule()
ligand = BSS.Parameters.parameterise(ligand_coord, 'gaff2').getMolecule()
system = protein + ligand

box = [100 * BSS.Units.Length.angstrom] * 3
angles = [100 * BSS.Units.Angle.degree] * 3
solvated = BSS.Solvent.solvate('tip3p', system, box=box, angles=angles, work_dir='tmp')

Expected behavior
BSS.Solvent.solvate() should not fail

(please complete the following information):

• Ubuntu
• Version of Python: 3.10.12
• Version of BioSimSpace: 2023.3.0
• I confirm that I have checked this bug still exists in the latest released version of BioSimSpace: [yes/no] no (not in the dev version)

[cespos]

I just noticed that this issue is related to issue #67

[lohedges]

Thanks. I am away at the moment and will take a look next week. We do have code to deal with existing waters, i.e. to make sure that they are contiguous, but it is clearly failing in certain cases.

[cespos]

Thanks Lester!

This solved it for me:

But maybe there is a better way to this internally/automatically

[lohedges]

Are you able to share the example system that triggers the error? I'm guessing that the issue is caused by one of the following:

• Non-standard water naming.
• Waters embedding in the protein, i.e. the are not separate molecules, so are not detected by our water finding code.

Cheers.

[cespos]

I cannot share the system but I can send you a screenshot of the PDB showing ions, water molecules, and the protein terminal residue.

[lohedges]

Thank you, this is very helpful. As shown, the first crystal water is combined with other atoms, i.e. MG, CA, and CA. We'll need to add some logic to search for water molecules within a structure, rather than just as self-contained molecules. I'll see if I can find another example, or make one myself.