AMBER Archive (2005)

Subject: RE: AMBER: min problem

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Fri Aug 19 2005 - 13:52:36 CDT


Dear Gustavo,
 
What you are seeing does is quite strange, often minimisation starts with
very high EEL or VDW terms that then shrink rapidly as you minimise. In your
case you see the opposite. However, I suspect that it is a problem with your
starting structure. Particularly since the electrostatics (EEL) are very
large and positive on step 50. What are they like on step 1?
 
Try checking the structure carefully, do you have any hydrogens pointing
through the middle of a ring etc. Also, before constructing the solvated
system try minimising in gas phase and then use this minimised gas phase
structure as the structure you choose to solvate. Also, is you system
charged? Have you added counter ions to try to neutralise the system? What
are you using for a cut off? If you are doing periodic boundaries with PME a
cut off of 8 angstroms is fine but without PME (gas phase, implicit solvent
or periodic boundaries with a regular cut off) cut should be set closer to
16 angstroms or so.
 
Note, the reason you see the LINMIN failures is because the RMS and gradient
maximum are huge >10^9 and so when you switch from steepest descent to
conjugate gradient it goes nuts since it tries to find the tangent of this
huge force vector and fails. For conjugate gradient to work you typically
need to be fairly close to the minimum.
 
All the best
Ross
/\
\/
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel: +1 858 784 8889 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk <http://www.rosswalker.co.uk/> | PGP Key
available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be ready every day, and should not be used for urgent or sensitive issues.

 

  _____

From: owner-amber_at_scripps.edu [mailto:owner-amber_at_scripps.edu] On Behalf Of
Gustavo Pierdominici Sottile
Sent: Tuesday, August 16, 2005 16:06
To: amber_at_scripps.edu
Subject: AMBER: min problem

Dear amber users: I have a problem in the minimization. If, when using leap
I do not solvate the system or if I use solvateCap, there is no problem. But
in case I use solvatebox and then perform a minimization, the output looks
like this:
  NSTEP ENERGY RMS GMAX NAME NUMBER
     50 5.0632E+05 5.7417E+00 2.3511E+01 O 9993
 BOND = 59276.8070 ANGLE = 332.6435 DIHED =
647.3963
 VDWAALS = 7513.5723 EEL = 436306.8605 HBOND =
0.0000
 1-4 VDW = 369.7308 1-4 EEL = 1869.0877 RESTRAINT =
0.0000

   NSTEP ENERGY RMS GMAX NAME NUMBER
    100 -3.7485E+08 9.5844E+09 1.1161E+12 H2 1865
 BOND = 67691.4229 ANGLE = 357.0795 DIHED =
650.5018
 VDWAALS = 11625.8570 EEL = ************* HBOND =
0.0000
 1-4 VDW = 375.1008 1-4 EEL = 1908.3052 RESTRAINT =
0.0000
 
There is a dramatical change in EEL, the RMS is very big, and the energy
decreases a lot. From this points up to the EEL =**************. There is
nothing rare with the atoms pointed in "NUMBER".
This seems to me very strange. The last structure seems to be Ok. There are
large fluctuations in the enegy values (e+08, e+11)and the message:
.... RESTARTED DUE TO LINMIN FAILURE ...
appears when the step number is the same as ncyc (when the method changes)
 
If someone note what is happening, please, let me know.
Thanking in advance

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu