[gmx-users] Adding TPO and SEP to G53a6 Forcefield

[Justin Lemkul]

On 2/26/14, 7:05 AM, lalithkumar wrote:
> Dear Dr. Justin,
>
> Thank you for pointing out the details of error. I have missed out the
> version of GROMACS I have been using and so I would like to let you know
> that my version is 4.5.7. In this version ffnonbonded.itp file of the
> gromos53a6.ff forcefield doesnt have pairtypes fr LJ-14 interactions. I

I doubt that's correct; every Gromacs version I've ever had contained 
[pairtypes] in this file.  Check again.  I don't have 4.5.7 installed anywhere, 
but in 4.6.x (several versions), it is clearly present.

> should agree that I dont know how they can be generated and to specify
> manually. So, in a crude way I can say that I have copied all the LJ 14
> types from the ffG43a1p forcefield at the gromacs contributors link:
> http://www.gromacs.org/Downloads/User_contributions/Force_fields
>
> This includes the following addition lines:
>
> ; begin insert from pamac_hsnnb.itp
> ; [ pairtypes ]
> ; for LJ 14 - use same for OP as for OA
>        OP        O  1  0.0022619536  9.687375e-07
>        OP       OM  1  0.0022619536  9.687375e-07
>        OP       OP  1  0.0022619536  1.265625e-06
>        OW       OP  1  0.0024331696   1.737e-06
>         N       OP  1  0.0023475616  1.463625e-06
>        NT       OP  1  0.0023475616  1.463625e-06
>        NL       OP  1  0.0023475616  1.463625e-06
>        NR       OP  1  0.0023475616  1.463625e-06
>        NZ       OP  1  0.0023475616  1.463625e-06
>        NE       OP  1  0.0023475616  1.463625e-06
>         C       OP  1  0.0023009528  2.066625e-06
>       CH1       OP  1  0.0025663376  2.174625e-06
>       CH2       OP  1  0.0032687988  3.000375e-06
>       CH3       OP  1  0.0039370168  3.907125e-06
>       CH4       OP  1  0.005459888  6.59475e-06
>       CR1       OP  1  0.003536086  3.24675e-06
>        HC       OP  1  0.000437552  1.38375e-07
>         H       OP  1           0           0
>       DUM       OP  1           0           0
>         S       OP  1  0.0047521952   4.068e-06
>      CU1+       OP  1  0.000972602  8.053875e-08
>      CU2+       OP  1  0.000972602  8.053875e-08
>        FE       OP  1           0           0
>      ZN2+       OP  1  0.000972602  1.09305e-07
>      MG2+       OP  1  0.0003842848  6.56775e-08
>      CA2+       OP  1  0.001507652  7.939125e-07
>         P       OP  1  0.005773784  5.299875e-06
>        AR       OP  1  0.003764374  3.53025e-06
>         F       OP  1  0.0016322592  9.81225e-07
>        CL       OP  1  0.0044525672  4.399875e-06
>        BR       OP  1  0.0016332104  9.1035e-06
>      CMET       OP  1  0.0044806276  5.1373125e-06
>      OMET       OP  1  0.0022619536  1.265625e-06
>       NA+       OP  1  0.00040373684  1.63125e-07
>       CL-       OP  1   0.0055883  1.16325e-05
>      CCHL       OP  1  0.0024394475   2.268e-06
>     CLCHL       OP  1  0.004334666  4.1738625e-06
>      HCHL       OP  1  0.0002920184  7.377075e-08
>     SDMSO       OP  1  0.0048877412  5.216175e-06
>     CDMSO       OP  1  0.0045248108  5.2475625e-06
>     ODMSO       OP  1  0.0022663291  9.752175e-07
>      CCL4       OP  1  0.0024394475  3.1014e-06
>     CLCL4       OP  1  0.0041472796  4.01985e-06
>        SI       OP  1           0           0
> ; end insert from pamac_hsnnb.itp
>
> And I had copied them as is to my ffnonbonded.itp file in modified
> gromos53a6.ff folder. Also, I have also modified atomtypes as per latest
> versions (for example CMET to CMet).
>
> Finally, I could pass the grompp set now and can run mdrun too successfully.
> Now I need your valuable review to know whether I had done is correct or can
> I do better than this way. If so please suggest some helping material
> regarding process of setting user defined LJ 14 pair types interactions to
> gromos53a6ff or any forcefield. And also, briefly comment on how difference
> with my results by doing in the method mentioned by me and with any other
> suggested method from your side.
>
> Please let me know if any further information to be provided from my side
>

What you've done is not correct.  You've made a mixed force field that is not 
necessarily viable.  You need proper parameters within a single force field, not 
borrowed from some other parameter set, unless you can demonstrate through 
rigorous testing and validation that what you've done is correct.

-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

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