Sigma value for alcohol hydrogen in OPLSAA xml

[mattiafelice-palermo]

Dear foyer developers,

I was building my own parameter xml file and copying the parameters from the oplsaa.xml file distributed with foyer, and I stumbled upon a possible inconsistency.

For the hydrogen type opls_155, which is described as "all-atom H(O): mono alcohols, OP(=O)2", the reported sigma is 1.0 Angstroms. I check on the original publication, and it seems that there they use a sigma value of 0.0:

I have also checked in the OPLSaa ffnonbonded.itp file provided by GROMACS and it also reports a value of 0 for sigma.

Even though this value does not make much difference since also epsilon is 0, it might make more sense to be consistent with the original publication and other sources?

Thank you for all the help and support!

Mattia

[daico007]

Hi @mattiafelice-palermo, that change of sigma value from 0 -> 1 was intentional. Some of the engine we support has to be normalized by sigma value, and having it as original value lead to dividing by 0 error. The modification should not affect the potential calculation/simulation since the epsilon of those atomtypes is also 0, which will cancel out the nonbonded force involving that atomtype (when plugged into the LJ equation below).