Of course it is possible but you need to check the free energy difference between the reactants and the products with free energy simulation. For instance, you can use metadynamics or umbrella sampling.

Thanks Zhaoxi

It is absolutely possible, yet It might be hard to model. Take into account that your cluster may not reproduce solvent effects, explicit hydrogen bonding or not, that might be highly relevant for the energy profile of the reaction. There is a sampling issue at hand that can surpass the purely reactive problem.

I interpret your post to mean that you have constructed a small molecule representation of the enzyme active site and are treating the molecules with some flavor of DFT. In my experience, such an approach will not yield reasonable results. The electrostatic environment within the active site significantly alters the free energy surface and small molecule analogs do not faithfully represent the enzyme. In my work, I chose to use QM/MM methods to retain the bulk of the enzyme through long-range electrostatic interactions. I would have preferred to just calculate everything QM but that was not tractable; QM/MM provided the only possible pathway at the time. These days, you might want to also consider ONIOM or other such layered quantum methods, depending on your computational resources. Ultimately you will need an estimate of the free energy profile along the reaction pathway. This will require repeated QM/MM calculations to provide statistics. The cheapest approach might be to use nudged elastic band (NEB) calculations to investigate the energy surface along putative pathways before committing to the more expensive free-energy sampling strategies.

NWChem provides a QM/MM method that supports DFT for the QM partition and Amber for the MM part. It is not user-friendly but the developers were quite helpful in providing guidance. Ruben is correct in his assessment that this will be difficult to model. It is possible, though. In several of my studies, water played the role of either donating or accepting a proton, on occasion through a chain of water molecules. Water is not just a mobile source of hydrogen bonds.

Thank you so much Ruben.

Hi Mark

Thanks for suggestion.

For DFT (gaussian) calculation, to consider rest of enzyme I am using eps=4