[gmx-users] Strange trjconv error

Justin Lemkul jalemkul at vt.edu
Tue Dec 10 16:00:37 CET 2013


On Tue, Dec 10, 2013 at 8:55 AM, rajat desikan <rajatdesikan at gmail.com>wrote:

> Hi All,
> I have a NPT NAMD trajectory of a membrane-protein that I want to analyze
> in gromacs 4.6.3. (g_density). I used catdcd to convert the .dcd to .pdb
> and generated a .top from the .psf using topotools in vmd. I then generated
> a .tpr using grompp
>
> When I do any simple gromacs command like:
>
> trjconv -f 130ns-141ns.pdb -s ref.tpr -o 130ns-141ns.xtc
>
> I get the following error:
>
> Fatal error:
> An input file contains a line longer than 4096 characters, while the buffer
> passed to fgets2 has size 4096. The line starts with: '20s'
>
>
Unfortunately this is an output bug, so the '20s' is not actually useful to
you.  Looks like something needs to be fixed in fgets2().


> wc -L ref.tpr
> >20344 ref.tpr
>
> The gmxdump of the .tpr along with grep gave me the offending line. It is
> essentially the protein description:
>
> atom[     0]={type=  0, typeB=  0, ptype=    Atom, m= 1.40067e+01,
> q=-6.00000e-01, mB= 1.40067e+01, qB=-6.00000e-01, resind=    0, atomnumber=
> -1}
>             atom[     1]={type=  0, typeB=  0, ptype=    Atom, m=
> 1.20107e+01, q=-1.00000e-01, mB= 1.20107e+01, qB=-1.00000e-01, resind=
> 0, atomnumber= -1}
>             atom[     2]={type=  0, typeB=  0, ptype=    Atom, m=
> 1.20107e+01, q=-3.50000e-01, mB= 1.20107e+01, qB=-3.50000e-01, resind=
> 0, atomnumber= -1}
>             atom[     3]={type=  0, typeB=  0, ptype=    Atom, m=
> 1.20107e+01, q=-3.50000e-01, mB= 1.20107e+01, qB=-3.50000e-01, resind=
> 0, atomnumber= -1}
>             atom[     4]={type=  0, typeB=  0, ptype=    Atom, m=
> 1.20107e+01, q=-3.50000e-01, mB= 1.20107e+01, qB=-3.50000e-01, resind=
> 0, atomnumber=
>
> -1}...........................................................................
>
> Any idea about how to proceed. I am quite stumped.
>
>
The problem is not in the .tpr file, it is in the .pdb file.  Something
about its format is bad.  Why not use catdcd to produce a .trr file instead?

-Justin

-- 

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

Justin A. Lemkul, Ph.D.
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


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


More information about the gromacs.org_gmx-users mailing list