[gmx-users] pdb vs xtc with g_bond and g_angle

Mark Abraham Mark.Abraham at anu.edu.au
Wed Dec 9 17:30:41 CET 2009


sandeep somani wrote:
> Hi 
> 
> I am trying to extract bond lengths and bond angles from a trajectory 
> file using g_bond and g_angle. 
> Original simulation generated a xtc file which I then converted to pdb 
> using trjconv. 
> 
> Turns out that for some frames bond lengths (and angles) computed using 
> the xtc file are different from that computed using pdb! And the pdb 
> values are correct as they match with measurements in VMD. Sample 
> outputs for the two cases are below. 
> 
> Any idea what is causing this discrepancy ? I am using GMX 4.0.5. 

Sounds like a periodicity artefact from your use of trjconv to me.

Mark

> *SAMPLE OUTPUT (distance.xvg) *
> *
> *
> matching values are in bold. 
> 
> -------------PDB------------------------
> $ g_bond_d -f propane4.pdb -n bond.ndx -s md.tpr
> 
> @ xaxis label "Time (ps)"
> @ yaxis label "Distance (nm)"
> *0.0000 0.15820*
> 1.0000 0.14859
> 2.0000 0.17248
> 3.0000 0.15750
> 4.0000 0.15835
> *5.0000 0.15367
> 6.0000 0.15394
> 7.0000 0.15878*
> 8.0000 0.15211
> 9.0000 0.15482
> 10.0000 0.14874
> 
> 
> -------------XTC------------------------
> $ g_bond_d -f propane4.xtc -n bond.ndx -s md.tpr
> 
> @ xaxis label "Time (ps)"
> @ yaxis label "Distance (nm)"
> *0.0000 0.15820*
> 1.0000 0.05791
> 2.0000 0.13811
> 3.0000 0.14364
> 4.0000 0.13509
> *5.0000 0.15367
> 6.0000 0.15394
> 7.0000 0.15878*
> 8.0000 0.06900
> 9.0000 0.07710
> 10.0000 0.08916
> 
> 
> 
> 
> 



More information about the gromacs.org_gmx-users mailing list