[gmx-users] Re: gmx-users Digest, Vol 53, Issue 75

Vitaly Chaban chaban at univer.kharkov.ua
Mon Sep 15 21:34:07 CEST 2008


> Vitaly Chaban wrote:
>> Hello,
>>
>> Trying to generate a topology from PDB file my x2top (gromacs 4.0beta) hangs up.
>> I have the coordinates of carbon nanotube in the PDB file (generated
>> with MOLEKEL). I also have
>>
>> ffencadv.n2t with:
>> C C 1 C ; CNT Carbon with one bond
>> C C 2 C C ; CNT double bonded Carbon
>>
>> ===============================
>> ffgmxbon.itp with:
>> [ bondtypes ]
>> ; i j func b0 kb
>> C C 1 0.14210 478900.
>> [ angletypes ]
>> ; i j k func th0 cth
>> C C C 1 120.000 397.480
>> [ dihedraltypes ]
>> ; i l func q0 cq
>> C C 1 0.000 167.360
>>
>> x2top -f cnt.pdb -o cnt.top -ff select
>> Then I select  "8: Encad all-atom force field, using scaled-down vacuum charges"
>>
>> Opening library file /usr/local/gromacs/share/gromacs/top/ffencadv.rtp
>> Opening library file /usr/local/gromacs/share/gromacs/top/ffencadv.n2t
>> Opening library file /usr/local/gromacs/share/gromacs/top/ffencadv.n2t
>> There are 0 name to type translations
>> Generating bonds from distances...
>> atom 0
>>
>> And the programme hangs up. The processor load is 99% all the time
>> before I terminate the job.
>>
>> The same data input works with GROMACS 3.3.1 as described by C. Stiles
>> but not with 3.3.3 and 4.0b however.
>> If it is not a bug then what should be modified to make a topology?
>> Has anyone tried to make SWCNT topology in the versions newer that
>> 3.3.1?
>>
>> What essential things have been changed in x2top since 3.3.1?
>>
>> Thanks a lot for suggestions!
>>
> The 3.3.3 and 4.0 versions are the same. Since I am busy completely
> reworking the code you can consider the current version broken. The
> topology from 3.3.1 will still work though...
>

The topology works but only in the case if all the parameters are
explicitly defined in the top file (x2top -param) generated in 3.3.1.


It seems the things are not so easy... In the force fields distributed
with gmx-3.3.3 there are no some n2t files which were present in
3.3.1. For what reason were they removed?

If I give to x2top the ffgmx.n2t (or ffencadv.n2t) from gmx-3.3.1 it
seems to read it
"
Opening library file /usr/local/gromacs/share/gromacs/top/ffencadv.rtp
Opening library file ffencadv.n2t
Opening library file ffencadv.n2t
"

but do not use. "There are 0 name to type translations"...
"
Generating bonds from distances...
atom 150
Can not find forcefield for atom C-1 with 0 bonds
Can not find forcefield for atom C-2 with 0 bonds
Can not find forcefield for atom C-3 with 0 bonds
Can not find forcefield for atom C-4 with 0 bonds
Can not find forcefield for atom C-5 with 0 bonds
"

The ffgmxbon.itp contains right parameters for bond distance, angles
and dihedrals. It is a great mystery for me why x2top does not work
correctly. And why carbon above has 0 bonds.

All the files are unix formatted with maximum rights of access...


Vitaly



More information about the gromacs.org_gmx-users mailing list