[AMBER] For Amber14 TI, how to perform only the discharge/recharge steps?

From: Lawrenz, Morgan <mlawrenz.amgen.com>
Date: Wed, 17 Sep 2014 01:01:04 +0000

Hi all,
I am trying to use Amber14 to perform TI and compute relative free energies between two molecules.
I'm working with a test system, two molecules in vacuum, and have reviewed the manual for the new single topology approach, along with recent tutorials I've successfully used before with the dual topology approach: http://ambermd.org/tutorials/advanced/tutorial9/setup.html

I've planned 3 steps, one to discharge molecule1, one to scale the vdw parameters from molecule 1 to molecule 2, and a third step to recharge molecule2.
I have step 2 running successfully (after turning off SHAKE) but have been unable to get step 1 going using the new flags. I am not sure if I understand the manual correctly.
For step 2, I made a topology with leap using a pdb of the two different molecules. For step 1, I thought that I needed to build a topology with a pdb file of two molecule1 structures (which are exactly the same), numbered residue 1 and 2 in the pdb file. The molecules have the same residue name. Then I specified the v0 and v1 atoms with timask1=":1" , timask2=":2" and then just use a charge mask for residue 2 (where I want partial charges gone).
I kept getting "Bus errors" and then turned off SHAKE and still am getting a segmentation fault when performing minimization.
Segmentation fault /common/compchem/src/amber14/bin/pmemd -O -i min-fep.in -o min-fep.out -p ../phenols.top -c ../phenols.crd -x min-fep.netcdf -r min-fep.rst

Here is my input file for minimization, please let me know if I am missing something!
minimization fep
  &cntrl
  imin = 1, ntmin=2, ntx = 1,
  maxcyc=500,
  ntpr = 100,
  ntf = 1, ntc = 1,
  ntb = 1, cut = 9.0,
  icfe=1,
  clambda=0.1,
  klambda=1,
  tishake=1,
  idecomp=0,
  timask1=":1",
  timask2=":2",
  ifsc=1,
  scalpha=0.5,
  scmask1=":1",
  scmask2=":2",
  logdvdl=1,
  crgmask=":2"
/

Thanks a lot for the help! And let me know if any more info is helpful here.
Morgan
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Sep 16 2014 - 18:30:02 PDT
Custom Search