Is Waterkit suitable for ligand bound system?

[zhentg]

Hi,

In the paper about Waterkit, at the "Structure preparation" section, it reads
"Crystallographic water molecules and Amber nonstandard residues were removed from the receptor, while ions were kept."

• Does that mean that ligand is also removed before Waterkit calculation?
• Why ligand is removed?
• Does Waterkit parameters work well with ligand (organic compound)?

Best
Zhenting

[jeeberhardt]

Hi Zhenting,

• Yes, in the paper all the ligands were removed before the calculations.
• We removed ligands because we wanted to study water molecule networks in the ligand binding site, and see how identified hydration sites were overlapping with hydrogen bond donor/acceptor features found in ligands.
• Yes, Waterkit calculations can be run in presence of ligands. Briefly, you will need to provide the GAFF2 parameters file (ligand.lib and ligand.frcmod) describing your ligand of interest at the PDBQT preparation step, and when you are generating the trajectory. See related issue Issue creating pdbqt file with ligand bound system #5 for more information.

Preparation step:

wk_prepare_receptor.py -i protein_ligand.pdb --lib ligand.lib --frcmod ligand.frcmod -o protein_prepared_ligand_test --pdb --amber_pdbqt

Trajectory generation step:

wk_make_trajectory.py -r protein_prepared.pdb -w traj -o protein --lib ligand.lib --frcmod ligand.frcmod

Hope it helps!

Best,

[zhentg]

Thanks for the answer. In fact I do run Waterkit with ligand successfully.
Hydration site analysis with ligand is very important for the affinity optimization of the ligand.
However, As the 1st step of Waterkit is to palce water around the solute, I wonder if current parameters are good enough to describe the small molecule hydration state?
I see in the supplementary table 1, there are SMARTS pattern defined for ligand hbond site.

Best
Zhenting

[jeeberhardt]

Well, that's a good question! I actually don't know if it works as well on protein-ligand complexes, since we never benchmark it for such applications. So that's something that it needs to be tested.

[zhentg]

I am doing tests with protein ligand complexes. Maybe we can continue to discuss this topic, and share the result. Jérôme Eberhardt ***@***.***> 于2023年8月28日周一 04:17写道：

…

Well, that's a good question! I actually don't know if it works as well on protein-ligand complexes, since we never benchmark it for such applications. So that's something that it needs to be tested. — Reply to this email directly, view it on GitHub <#7 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJVMMIJQHIUJOQHAEXPE6ALXXOTMXANCNFSM6AAAAAA36EI4ZU> . You are receiving this because you authored the thread.Message ID: ***@***.***>