Hi Maksim,
In this case the manual is correct and the code is wrong. The factor of 0.5 is need for the internal energy of the total system or some component of the system (e.g., the solute energy) to account for double counting. For the solute-solvent energy, there is no double counting and the factor should be omitted. For a relevant reference, see
H.-A. Yu and M. Karplus, The Journal of Chemical Physics 89, 2366 (1988).
I’ll attempt to get a patch out soon. In the mean time, please delete the factor in the code and recompile.
Thank you for finding and reporting this error.
Tyler
> On Sep 14, 2016, at 12:16 PM, Maksim Mišin <maksim.misin.strath.ac.uk> wrote:
>
> Dear RISM developers,
>
>
> I believe the 3D-RISM code in AmberTools outputs incorrect values of
> solute-solvent potential energy. The code (function
> rism3d_closure_solvPotEne in rism3d_closure_c.F90) produces a value that
> corresponds to:
> E_uv = 0.5*sum[ g(r)*u(r) ]*dr,
> while the Amber16 manual (p. 103) states that it should output:
> E_uv = sum[ g(r)*u(r) ]*dr,
> which is the correct equation for solute potential energy.
>
>
> Kind regards,
> Maksim Mišin
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Sep 14 2016 - 15:30:04 PDT
