EDISPER = 0 with RSFF2C force field: Silent fallback to inp=1?

[Umut Ç.]

Hi,

I performed MM/PBSA calculations using gmx_MMPBSA with the RSFF2C force field. However, I did not set inp, so it should default to 2.

&pb indi=4.0, istrng=0.0, /

EDISPER = 0.00 in Complex, Receptor, Ligand, and Delta across all frames. 

ENPOLAR is non-zero (~+71 kcal/mol for complex, ~+68 for receptor, ~+9 for ligand, ~−6 for delta). 

Did the calculation silently fall back to inp=1 because RSFF2C atom types aren't recognized for the Tan/Luo dispersion parameterization? Should non-Amber force fields always explicitly set inp=1? Also, is there a way to verify which inp value was actually used from the output files? None of the output files report the used default parameter settings.

Thanks.

[marioe911116]

Double-check the _GMXMMPBSA_info file and look for INPUT['pb']['inp']. In any case, specify the inp value in the input file. 

&pb 

 indi=4.0, 

 istrng=0.0, 

 inp=2, 

/

It seems there is a mismatch between the docs and the default value for inp variable. We are going to fix this in teh coming version.

[Umut Ç.]

I see. I couldn't recover the _GMXMMPBSA_info file from my actual runs since I used the keyword keep_files = 0 in my input file, but I was lucky enough to recover it in my trial run folder. It might be a good addition to keep this file even when keep_files = 0 is used in the input file for future reference.  In the _GMXMMPBSA_info file, I have:

INPUT['pb']['inp'] = 1

which means that it automatically falls back. Thank you for your quick and helpful response.