[gmx-users] gromos function type assigned by grompp

[Justin A. Lemkul]

Romelia Salomon wrote:
> Hi
> 
> I am trying to set a calculation based on the GROMOS force field in a
> program I have and I am using GROMACS to generate the (starting) input
> files for me. From reading the manual I know that GROMOS uses a fourth
> power potential (type 2) for bonds and a cosine based angle potential
> (type 2), but I noticed that when using grompp to include the values for
> the parameters explicitly it changes the type from 2 to 1, for example:
> 
> topology file before grompp
> [ bonds ]
> ;  ai    aj funct            c0            c1            c2            c3
>     1     2     2    gb_2
>     1     3     2    gb_2
>     1     4     2    gb_2
>     1     5     2    gb_20
>     5     6     2    gb_26
>     5    10     2    gb_26
> after using grompp

This would be from the -pp flag of grompp?

> [ bonds ]
> ;  ai    aj  tp parameters
>     N    H1   1       0.1 1.87e+07
>     N    H2   1       0.1 1.87e+07
>     N    H3   1       0.1 1.87e+07
>     N    CA   1     0.147 8.71e+06
>    CA    CB   1     0.153 7.15e+06
>    CA     C   1     0.153 7.15e+06
> 

Which version of Gromacs are you using?  I'm using 3.3.3, and mine works just 
fine (bond type is 2).  Furthermore, it's odd that grompp is including atom 
names instead of numbers...unless you've interpreted what the output really is 
and transcribed it differently :)

If grompp is changing these, then perhaps it's a bug.  Note that weird behavior 
arises when Gromacs is compiled with gcc 4.1.x, as has been beaten to death on 
this list :)  Without knowing more about your system, this is pure speculation 
and may not be worth anything.

> Is this right? Also, for clarification, I know Gromacs uses kj/mol and nm
> as units, so does that mean that the values included in the ff*.itp files
> under the share/top/ directory are in those units?
> 

The units are all consistent.

-Justin

> Also I noticed that the .itp files for bonding interactions in gmxbon.itp
> have declared a type 1 function which would be the regular harmonic one.
> 
> Sorry for the very basic questions, but I wanted to be certain about this
> things, thanks!!
> 
> Romelia
> 
> Thanks!
> 
> 

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

Justin A. Lemkul
Graduate Research Assistant
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

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