Re: Multiplicity of Parameters

From: Philip Peartree (P.Peartree_at_postgrad.manchester.ac.uk)
Date: Fri Aug 10 2007 - 07:25:33 CDT

Thanks Peter, I found my problem (although not through the psf) I was looking
through the topology file, and in a custom patch statement (used to construct
the modified residue) a dihedral had been created twice. Thanks for the help,
it's now working fine (fingers crossed!)

Philip Peartree

Quoting Peter Freddolino <petefred_at_ks.uiuc.edu>:

> Hi Philip,
> sorry, I had it muddled in my own head; the usage in the topology files
> is correct, and as long as you're using it the same way, things should
> work fine. Please note that in psfgen DOUBLE is treated exactly like
> BOND, since any differences are taken care of by the parameters. It
> would be good to check your psf, though, and see if any of the dihedrals
> are double counted. I believe this could happen, for example, if you
> both explicitly specify a dihedral and indicate autogeneration (although
> I'd have to test that to be sure -- I know I've seen this error show up
> before when I did something along those lines).
> Best,
> Peter
>
> Philip Peartree wrote:
>> Hi Peter,
>>
>> I'm not sure what you mean in your last sentence, as far as I can see,
>> in all
>> residues in the topology file double is used in place of the bond
>> keyword to
>> specify a double bond instead of a single bond. Is this not correct
>> usage?
>>
>> Unfortunately I don't have the psf at hand, but I will check it tomorrow.
>>
>> Philip
>>
>>
>> Quoting Peter Freddolino <petefred_at_ks.uiuc.edu>:
>>
>>> Hi Philip,
>>> I honestly don't know if charmm does any sort of internal checking to
>>> avoid this being a problem. I can suggest looking in your psf file at
>>> any dihedral terms involving four CA atoms, and see if any of them are
>>> listed repeatedly. If so, you may want to make sure that in your
>>> topology file no bonds are doubled up, and that you're not applying a
>>> patch that lists an already existing bond. Also, you should make sure
>>> that the double keyword is not being used in the topology file without
>>> also specifying a bond for the same two atoms; iirc correctly psfgen
>>> doesn't like this.
>>> Best,
>>> Peter
>>>
>>> Philip Peartree wrote:
>>>> Hi Peter,
>>>>
>>>> I generated the psf using the VMD psf gen tool (the graphical one, not
>>>> the
>>>> commandline version) and the topology file is one of my own making
>>>> (essentially
>>>> Charmm27 but with a few additional residues (I'm simulating a
>>>> quinoprotein,
>>>> which contains a modified tryptophan) I'm assuming that the problem is
>>>> with
>>>> that residue, but the thing puzzling me is that it worked fine in c31
>>>> itself.
>>>>
>>>> Philip Peartree
>>>>
>>>>
>>>>
>>>> Quoting Peter Freddolino <petefred_at_ks.uiuc.edu>:
>>>>
>>>>> Hi Philip,
>>>>> just to verify, could you please let us know how you generated your
>>>>> psf?
>>>>> And which topology file are you using?
>>>>> Best,
>>>>> Peter
>>>>>
>>>>> Philip Peartree wrote:
>>>>>> Dear All
>>>>>>
>>>>>> Upon running (an albeit) basic simulation (Minimisation, I
>>>>>> received the
>>>>>> following error:
>>>>>>
>>>>>> FATAL ERROR: Multiplicity of Paramters for diehedral bond CA CA CA
>>>>>> CA of 1
>>>>>> exceeded
>>>>>>
>>>>>> I am aware that several messages of this type have been posted to
>>>>>> this list, but
>>>>>> none have fully detailed a positive outcome. From what I have read
>>>>>> of other
>>>>>> messages, this is something to do with a wrongly typed bond
>>>>>> somewhere in the
>>>>>> topology file (I am using a Charmm type topology). However, this
>>>>>> topology/parameter set worked fine in Charmm31 (I have decided to
>>>>>> start running
>>>>>> in NAMD since it was taking about 3.5 months to run a 10 ns
>>>>>> simulation!!)
>>>>>>
>>>>>> What should I do about correcting this, since the bonds are
>>>>>> correctly defined in
>>>>>> the topology (as far as I know)
>>>>>>
>>>>>> Philip Peartree
>>>>>> Sutcliffe Group
>>>>>> University of Manchester
>>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:46:46 CST