[gmx-users] Re: No such moleculetype SOL

Justin Lemkul jalemkul at vt.edu
Mon Sep 23 04:18:09 CEST 2013



On 9/22/13 10:10 PM, Jonathan Saboury wrote:
>> Then something is wrong with the way you've done it, because the job of
> spc.itp
>> is to define the [moleculetype] SOL and all its relevant parameters.
>
> Hm...not sure what the matter is. Added a forcefield as well as spc.itp to
> the .top file made by acpype. Still getting the error "Fatal error: No such
> moleculetype SOL"
>
> Here is a list of commands and explanations of what I did:
> http://pastebin.com/raw.php?i=V5wrRdme
> Here is the .top: http://pastebin.com/raw.php?i=LzCANQZw
> Here are the files I am using: http://www.sendspace.com/file/5ps0i1
>

I should have seen this before.  The default spc.itp is not compatible with 
AMBER.  Have a look at the actual contents of spc.itp:

; This file is for backward compatibility only.
; Please directly include the spc.itp file from your force field directory.
#ifdef _FF_GROMACS
#include "gmx.ff/spc.itp"
#endif
#ifdef _FF_GROMOS96
#include "gromos43a1.ff/spc.itp"
#endif
#ifdef _FF_OPLS
#include "oplsaa.ff/spc.itp"
#endif

That said, there is an spc.itp within the AMBER subdirectories that needs to be 
#included more explicitly, i.e. #include "amber99sb.ff/spc.itp"

May I ask why you are using SPC?  The AMBER force fields were parametrized with 
TIP3P, so I see no viable reason to use a different water model.

> It is important to note that genbox did not add the amount of cyclohexanes
> added to system in the .top (it did do it for the waters though). Perhaps
> that is a bug?
>

It's not a bug, because genbox does not advertise such a feature.  genbox is 
hard-coded to only ever deal with water.  Anything else is up to the user.

>> You need a parent force field that defines all the bonded and nonbonded
>> interactions in the system.  Whether or not this is one that comes with
> Gromacs
>> or one you design yourself, you still have to satisfy all of the internal
>> mechanics (see Chapter 5).  Whether or not there is a protein is completely
>> irrelevant; force fields do much more than just proteins.
>
> I see, interesting. Added it to the .top :)
>
>> Well, PRODRG parameters are pretty low-quality anyway, but it's a
> convenient
>> example for doing a tutorial rather than trying to teach people how to use
> other
>> external software ;)
>
> As you know I am using acpype, is it any better? What do you use?
>

Any force field can be parametrized to give correct target data, especially for 
a molecule as simple as cyclohexane.  What I use is a complex question, because 
it depends entirely upon the task at hand :)

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

==================================================



More information about the gromacs.org_gmx-users mailing list