Non-optimal GB parameters

[Aron Broom]

Hi All,

I'm getting a user warning about non-optimal GB parameters when trying to run an implicit solvent simulation from LEaP inputs. Based on the code generating the warning this is because of an apparent mismatch between the radii and the screening terms.

Specifically, I'm selecting the "GBn2" model in OpenMM, and my LEaP inputs were generated by first setting the correct radii using: "set default PBRadii mbondi3" as indicated in the AMBER manual.

So as far as I can tell everything should match up.

I tried modifying the code in the amber_file_parser.py to take the screening terms from the prmtop even when using the GBn2 model and not just for the HCT and OBC models, but the same warning appears.

Has anyone else seen this? Is this an issue with generating the inputs using LEaP instead of loading the AMBER FF using OpenMM itself?

Thanks,

Aron

[Peter Eastman]

I think you can ignore that warning. It's checking for the mbondi2 radii. Let me check with the person who originally wrote that code to be sure.

[Aron Broom]

OK thanks.

The simulation qualitatively looks OK, nothing bizarre happening.