Re: [AMBER] Equilibration of membrane protein

From: Benjamin D Madej <>
Date: Tue, 10 Jun 2014 19:04:41 +0000

Hi James,

One noticeable thing about your input is that the nstlim and istep2 do not match. I would guess that you're heating your system too fast, leading to instabilities. It's worth checking the input structure as well for any irregularities.

The error you show is a GPU specific error. Only pmemd.cuda could give you that error. I would attempt to run the simulation again with just "pmemd" and it's likely that you'll get more information about what's going wrong with your simulation.

All the best,
Ben Madej
UCSD Chemistry and Biochemistry

From: James Starlight []
Sent: Tuesday, June 10, 2014 12:05 AM
To: AMBER Mailing List
Subject: [AMBER] Equilibration of membrane protein

Dear Amber users!

I've faced with the problem of the initial heating of my protein-membrane
system in NVT using script from the tutorial (before this stage my protein
have been properly minimized):

   9600 -2.0119E+05 7.8407E-01 7.5497E+01 C112 21832

 BOND = 16716.8403 ANGLE = 3296.3449 DIHED =
 VDWAALS = 25033.9597 EEL = -277016.9354 HBOND =
 1-4 VDW = 3857.1824 1-4 EEL = 14205.2856 RESTRAINT =

the nvt script is

Lipid 128 heating 100K
  value2=100.0 /
 &wt type='END' /
Hold lipid and protein fixed
RES 1 754 # <<< restrains applied both on protein and membrane residues

the error (occurred both in pmemd and sander) is

 pmemd.cuda -O -i ./in/ -o 02_heat.out -p test.prmtop -c
01_Min.rst -r 02_Heat.rst -ref 01_Min.rst -x
Error: unspecified launch failure launching kernel kClearForces
cudaFree GpuBuffer::Deallocate failed unspecified launch failure
AMBER mailing list

AMBER mailing list
Received on Tue Jun 10 2014 - 12:30:03 PDT
Custom Search