Re: [AMBER] Problem In Performing MMPBSA Calculation for System Containing Lipids

From: Jason Swails <>
Date: Wed, 26 Aug 2015 08:41:47 -0400

On Tue, Aug 25, 2015 at 8:04 PM, <>

> Hi Dan,
> I checked the trajectory through the cpptraj 'check' command and found no
> issue. I tried to run with the serial version also, it is giving the same
> error as printed bellow.
> File "/home/anilkumarsahoo/amber14/bin/", line 104, in ?
> app.parse_output_files()
> File "/home/anilkumarsahoo/amber14/bin/MMPBSA_mods/", line 928,
> in parse_output_files
> self.using_chamber)}
> File "/home/anilkumarsahoo/amber14/bin/MMPBSA_mods/",
> line 634, in __init__
> AmberOutput._read(self)
> File "/home/anilkumarsahoo/amber14/bin/MMPBSA_mods/",
> line 343, in _read
> self._get_energies(output_file)
> File "/home/anilkumarsahoo/amber14/bin/MMPBSA_mods/",
> line 661, in _get_energies
> ValueError: invalid literal for float(): *************
> Exiting. All files have been retained.
> Any help will be really appreciated.

​To add a little to Dave's advice, this often occurs when the solvated
prmtop without the atoms defined in strip_mask *doesn't* match the complex
prmtop you are using. It can also happen when the ligand mask doesn't
match the ligand component of the complex prmtop.

If you are using the CHARMM force field through the 'chamber' program (or
the ParmEd 'chamber' action), then the water and ions have different names
than Amber uses (TIP3 for water, SOD for sodium, and CLA for chloride, for
example). You will need to set the "strip_mask" variable in the &general
section of your input file to remove the necessary atoms from the solvated

Another comment I'll make is that even after you do this, the only mode
that really supports for doing membrane binding energies is to
consider the lipids explicitly (there are no membrane GB models in Amber,
and doesn't support the PB membrane options). There are better
approaches to computing reaction fields in membrane environments using PB,
and you will likely have to carry those out by hand using either the
membrane options in PBSA or those in APBS. This is, of course, assuming
that your lipids are in a membrane and not a single one bound in some way
to a protein.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Wed Aug 26 2015 - 06:00:06 PDT
Custom Search