AMBER Archive (2008)

Subject: Re: AMBER: interaction energies in Amber and VMD/NAMD

From: Alan (alanwilter_at_gmail.com)
Date: Tue Jul 29 2008 - 05:10:48 CDT

Alessandro,

I am reading again Cornell et all. 1994 and found out that for ele 1-4
factor is 1/1.2 and for vdw 1-4 it is 1/1.2. It occurred to me now
that NAMD may use a unique factor (1/1.2) if running amber input files
in NAMD.

I acknowledge that I may be being confused here, but if you have time
why not test your simulation in NAMD with '1-4scaling 0.5'? My guess
is that it would reproduce your sander vdw results but then failed for
ele.

Just thinking aloud.

Cheers,
Alan

On Tue, Jul 29, 2008 at 9:19 AM, Alan <alanwilter_at_gmail.com> wrote:
> Looking only at the figure, it seems there's a factor of 2x in amber
> vdw x namd vdw. I really don't know which one is correct and I don't
> know how Namd plugin is working.
>
> However you said you didn't put 1-4vdw term in your program, right?
> Well, I would say that it would do a good difference in your results.
>
> Is it possible for your to send me a spreadsheet with vdw values from
> amber and namd?
>
> Have you tried with switching 'on'?
>
> Sorry for asking again because I may missing something else here: so
> your ran a simulation in sander and namd2 right? and now you are
> trying to compare ele and vdw results, right?
>
> Which traj files are you trying with VMD NAMDEnergy? AFAIK VMD can
> read sander traj files and you can see results with NAMDEnergy. Are
> you trying to compare sander x namd2 or just checking if your program
> is working fine for sander as NAMDEnergy seems to work for namd2?
>
> Cheers,
> Alan
>
> On Mon, Jul 28, 2008 at 1:38 AM, Alessandro Nascimento
> <al.s.nascimento_at_gmail.com> wrote:
>> Thanks, David, Carlos and Alan.
>>
>> I am attaching a picture showing how different the results are.
>>
>> These results were from a simulation of a waterbox (TIP3P) using
>> sander. I choose a waterbox to debug the contribution of the 1-4vdw
>> term (which is missing so far in my simple code).
>>
>> The system was setup using leap as follows:
>>
>> wat = createunit wat
>> solvatebox water TIP3PBOX 12.0
>>
>> Then I minimized the system in sander, and ran a productive run
>> (100000 steps) also in sander, turning ntb=0.
>>
>> So far, I ve being defining vdw as :
>>
>> iaci = Natomtypes*(iac[i]-1);
>> ic = ico[(iaci+iac[j])-1]; // the `-1` corrects
>> for the array index which start in 0 rather than 1
>> if (ic > 0) {
>> vdw = vdw + (((LJA[ic-1])/(r2*r2*r2*r2*r2*r2))-((LJB[ic])/(r2*r2*r2))); }
>> else {
>> vdw = vdw +
>> (((LJA[ic-1])/(r2*r2*r2*r2*r2*r2))-((LJB[ic])/(r2*r2*r2*r2*r2))); }
>>
>> Using vmd to compute the interaction energy between molecule 1 and the
>> other ones (using the parameters suggested in the namd manual for
>> amber ff simulations:
>>
>> switching off
>> exclude 1-4scaling
>> 1-4scaling 0.83333
>> cutoff 35.0
>>
>> gives the results I show you in the picture
>>
>> Any idea of what I may be missing?
>>
>> Thanks a lot,
>>
>> --alessandro
>>
>>
>>
>> On Sat, Jul 26, 2008 at 11:26 PM, Alan <alanwilter_at_gmail.com> wrote:
>>> Alessandro,
>>>
>>> Have you looked at http://www.ks.uiuc.edu/Research/namd/2.6/ug/node15.html?
>>>
>>> Pay well attention to:
>>>
>>> 3. NAMD has several exclusion policy options, defined by exclude. The
>>> way AMBER dealing with exclusions corresponds to the ``scaled1-4'' in
>>> NAMD. So for simulations using AMBER force field, one would specify
>>> ``exclude scaled1-4'' in the configuration file, and set 1-4scaling to
>>> the inverse value of SCEE as would be used in AMBER.
>>>
>>> 5. By default, NAMD applies switching functions to the non-bond
>>> interactions within the cutoff distance, which helps to improve energy
>>> conservation, while AMBER does not use switching functions so it
>>> simply truncates the interactions at cutoff. However, if ``authentic''
>>> AMBER cutoff simulations are desired, the switching functions could be
>>> turned off by specifying ``switching off'' in NAMD configuration file.
>>>
>>> and check those parameters in your namd.conf file:
>>> 1-4scaling 0.833333 # =1/1.2, default is 1.0
>>> scnb 2 # This is default
>>> switching off # Turn off the switching functions, try on also
>>> readexclusions yes #you may compare with 'no' too.
>>>
>>> And , as David ask, how much is the difference in percentage?
>>>
>>> Cheers,
>>> Alan
>>>
>>> On Sun, Jul 27, 2008 at 1:46 AM, Carlos Simmerling
>>> <carlos.simmerling_at_gmail.com> wrote:
>>>> I don't think Alessandro used Amber for energies, and if it's a simple program
>>>> it probably is non-periodic. I think it's likely the 1-4 scaling factor for vdw.
>>>> one of them ( can't recall, Adrian may know) either ele or vdw has the
>>>> wrong 1-4 scaling in NAMD for amber ff.
>>>>
>>>> On Sat, Jul 26, 2008 at 8:36 PM, David A. Case <case_at_scripps.edu> wrote:
>>>>> On Sat, Jul 26, 2008, Alessandro Nascimento wrote:
>>>>>>
>>>>>> I wrote a simple program to read some information in the prmtop file
>>>>>> and compute nonbonded energies (ele+vdw) from a sander/pmemd
>>>>>> trajectory and I am trying now to compare my first test results with
>>>>>> the ones from the plugin NAMDEnergy in VMD. For electrostatic
>>>>>> computation, they match perfectly. For vdw, however, seems like
>>>>>> vmd/namd overestimates the energetic contribution. Does anyone has any
>>>>>> experience on comparing results between amber and vmd/namd ?
>>>>>
>>>>> I don't have experience, but as a guess: Amber by default includes a
>>>>> correction for vdw interactions beyond the cutoff; NAMD might not inlcude
>>>>> such a term. How different are the results?
>>>>>
>>>>> ...dac
>>>>>
>>>>> -----------------------------------------------------------------------
>>>>> The AMBER Mail Reflector
>>>>> To post, send mail to amber_at_scripps.edu
>>>>> To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
>>>>> to majordomo_at_scripps.edu
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> ===================================================================
>>>> Carlos L. Simmerling, Ph.D.
>>>> Associate Professor Phone: (631) 632-1336
>>>> Center for Structural Biology Fax: (631) 632-1555
>>>> CMM Bldg, Room G80
>>>> Stony Brook University E-mail: carlos.simmerling_at_gmail.com
>>>> Stony Brook, NY 11794-5115 Web: http://comp.chem.sunysb.edu
>>>> ===================================================================
>>>> -----------------------------------------------------------------------
>>>> The AMBER Mail Reflector
>>>> To post, send mail to amber_at_scripps.edu
>>>> To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
>>>> to majordomo_at_scripps.edu
>>>>
>>>
>>>
>>>
>>> --
>>> Alan Wilter S. da Silva, D.Sc. - CCPN Research Associate
>>> Department of Biochemistry, University of Cambridge.
>>> 80 Tennis Court Road, Cambridge CB2 1GA, UK.
>>>>>http://www.bio.cam.ac.uk/~awd28<<
>>> -----------------------------------------------------------------------
>>> The AMBER Mail Reflector
>>> To post, send mail to amber_at_scripps.edu
>>> To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
>>> to majordomo_at_scripps.edu
>>>
>>
>>
>>
>> --
>> [ ]s
>>
>> --alessandro
>>
>
>
>
> --
> Alan Wilter S. da Silva, D.Sc. - CCPN Research Associate
> Department of Biochemistry, University of Cambridge.
> 80 Tennis Court Road, Cambridge CB2 1GA, UK.
>>>http://www.bio.cam.ac.uk/~awd28<<
> 

-- 
Alan Wilter S. da Silva, D.Sc. - CCPN Research Associate
Department of Biochemistry, University of Cambridge.
80 Tennis Court Road, Cambridge CB2 1GA, UK.
>>http://www.bio.cam.ac.uk/~awd28<<
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo_at_scripps.edu