I recently installed the 2026.1 version.
Initially, I chose opc water model, but I couldn’t find any opc.gro, so I used tip4p.gro. However, it occasionally returned a “Not enough replaceable solvent molecules”.
Clearly, tip4p and opc are fundamentally different. So, I used the geometry info available in opc.itp to create a new opc.gro. The steps are as follows:
- Keep all the OW coordinate in
tip4p.grounchange. –> Fine-tune the coordinate of hydrogens using the correct bond lengths and angles using minimum displacement. –> Recalculate the coordinate of all virtual sites. - Remove all velocity data of the original
tip4p.gro - The cubic box margin seems to be the best at 0 nm.
- Energy minimization –> NVT –> NPT for 1 ns.
The final opc.gro and the preparation processes are posted on my Github.
I have used this to run my mdrun using amber19sb for 12 proteins. So far, it was smooth, no error.
If you are interested, please consider have a play with it.
Inputs and comments are welcomed!