Re: [AMBER] could not find vdW (Na+) type IP

From: Jason Swails <jason.swails.gmail.com>
Date: Mon, 12 Jan 2015 11:31:50 -0500

> On Jan 12, 2015, at 10:52 AM, Francesco Pietra <chiendarret.gmail.com> wrote:
>
> I tried to follow correctly your kind suggestion
>
> In leap:
> parm10 = loadamberparams parm10.dat
> frcmod99SB = loadamberparams frcmod.ff99SB
> frcmodionsff99_tip3p = loadamberparams frcmod.ionsff99_tip3p
> frcmodzn = loadamberparams frcmod.MOL.known
> frcmodzn = loadamberparams frcmod.MOL.correspondence
> frcmodzn = loadamberparams frcmod.added_impr
> loadOff all_amino94.lib
> loadOff all_aminoct94.lib
> loadOff all_aminont94.lib
> loadOff ions94.lib
> loadOff solvents.lib

Is there a reason you are trying to reinvent what the leaprc files do? I would not suggest the current strategy that you are employing, since it is way too easy to make a mistake. For instance -- parm10.dat contains the frcmod.ff99SB modifications already, but I think some of the side-chain atom types for the amino acids may be slightly different (in preparation for the modifications that became ff12SB and ff14SB).

Furthermore, the atom types used in parm10.dat correspond to the atom typing used in amino10.lib, aminoct10.lib, and aminont10.lib, but you are using the library files that were used for ff94 through ff99SB (excluding ff03). Lastly -- and what is causing your error here -- you are using the “old” definition of the ion residues/atom types (ions94.lib), whereas frcmod.ionsff99_tip3p is designed to be used with atomic_ions.lib, if I’m not mistaken.

A lot of work has gone into refining the ion parameters, so all of the atom types for these ions have been completely redone in a more sensible way... but what this means is that the old library files and “new” frcmod files are completely incompatible. The frcmod.ionsff99_tip3p file seems to have been provided in order to allow the “old” (i.e., ff99-and-older) ion parameters to be used with the “new” library files.

The takeaway message here is: use the leaprc files that have been prepared for you. Mixing and matching like you are doing here rarely results in desired behavior, and I would discourage anybody from ever doing this.

Good luck,
Jason

--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jan 12 2015 - 09:00:03 PST
Custom Search