[gmx-users] Non-bonded interactions

[Justin Lemkul]

On 6/4/14, 7:27 AM, 라지브간디 wrote:
>
> Dear justin,
>
> Shall I introduce this calculated non-bonded interaction values in gromos43a1.ff ( ffninbonded.itp) section or
>
> specifically I can include them in my protein topology file ? ([nonbond_param]) ?
>

Neither.  I have said it twice before, and I'll say it a third time - the 
parameters you have cannot be sensibly combined with others in the Gromos96 43A1 
parameter set.  They were developed for CHARMM; they need to be used with a 
CHARMM force field.  Hacking them into some other force field breaks the 
internal self-consistency of the force field and you end up with a junk simulation.

The proper approach is to introduce the parameters in [atomtypes] of 
ffnonbonded.itp in the CHARMM force field and let the combination rules do the 
rest of the work.

-Justin

>
>
> On 6/3/14, 4:56 AM, ????? wrote:
>> Dear Justin,
>>
>>
>> I have calculated upon the way you provided as follows :
>>
>>
>> CHARMM Rmin/2 -> Gromacs sigma:
>> multiply by 2 to get Rmin, convert to sigma by
>>
>> multiplying by 2^(-1/6), then multiply by 0.1 to convert A -> nm
>>
>>
>> Sigma = 3.83
>> Rmin = (3.83 * 2)
>>
>>
>> 7.66 * 2^(-1/6)*0.1 = 0.682428418
>>
>>
>> Epsilon: multiply the CHARMM value by 4.184 to convert kcal/mol to kJ/mol
>>
>>
>> Epsilon 0.0262 x 4.184 = 0.1096208
>>
>>
>> Is that right conversion?, if so, I have used g_sigeps utility in gromacs for above mentioned sigma and epsion values and got entirely different values:
>>
>>
>> g_sigeps -sig 3.83 -eps 0.0262
>>
>>
>> c6 = 3.30791e+02, c12 = 1.04411e+06
>> sigma = 3.83000, epsilon = 0.02620
>> Van der Waals minimum at 4.29903, V = -0.0262
>>
>>
>> Fit of Lennard Jones (12-6) to Buckingham:
>>
>> A = 4264.18, B = 2.79133, C = 330.791
>>
>>
>> Please advice if i am making any mistake ?
>>
>
> You're not supplying the right values to g_sigeps. It's designed to convert
> known values of sigma and epsilon (already in Gromacs units) to C6/C12, which is
> not what you need to accomplish. The value you're passing to -sig is actually
> Rmin, not sigma, and the value you're supplying to -eps is in the wrong units.
>
> The converted values you obtained with by-hand calculation (not g_sigeps) appear
> reasonable. Verifying the energy reported by CHARMM/NAMD and Gromacs should be
> a simple test to confirm.
>
> -Justin
>

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA 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
http://mackerell.umaryland.edu/~jalemkul

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