Re: [AMBER] Issues with running simulation after ParmEd conversion

From: Jason Swails <>
Date: Thu, 30 Jul 2015 16:53:14 -0400

On Thu, Jul 30, 2015 at 4:06 PM, Rohith Mohan <> wrote:

> I didn't see a warning about ATOMS_PER_MOLECULE so I think the PSF and
> Amber files should be compatible. I generated a chamber-free topology using
> cpptraj and visualizing in VMD still resulted in a warped image.
> While comparing the coordinates from the inpcrd file and the NAMD PDB file
> I noticed that for several coordinates there was a mismatch in the sign of
> a coordinate. For example, -23.667 -13.159-150.455 from the PDB becomes
> -23.6670000 -13.1590000 150.4550000. -150.455 becomes 150.455 which results
> in the structure being warped I believe. Additionally in some cases there's
> a lack of space between coordinates so I wonder if that may affecting the
> structure as well when AMBER reads it e.g: -95.272-102.260-121.094 from the
> PDB becomes -95.2720000-102.2600000 121.0940000 in the .inpcrd file.

Yep, I can confirm this. The problem is that the PDB parser skipped the
first character in the Z-coordinate field. I never saw it before because
99.9% of all Z-coordinates in PDB files are 7 fields or fewer, so it
usually just skipped whitespace. In your case, it neglected a negative
sign and completely wrecked the resulting inpcrd file.

Luckily the fix is exceedingly easy -- I will make an update for AmberTools
15 shortly. In the meantime, you can make the fix by hand. Go to the file
$AMBERHOME/AmberTools/src/parmed/chemistry/formats/ and look for line
194. It should read:

x, y, z = line[30:38], line[38:46], line[47:54]

Change the 47 to a 46 and make sure you recompile. To do this just for
ParmEd, you can go to $AMBERHOME/AmberTools/src/parmed and type:
​​python install --prefix=$AMBERHOME
Thanks for the report, and let me know if you still have problems,

Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Thu Jul 30 2015 - 14:00:02 PDT
Custom Search