amber-developers: Problems with restrts and igb>0

From: Ross Walker <ross.rosswalker.co.uk>
Date: Sat, 30 Oct 2004 19:40:06 -0700

Hi All,

Has anybody come across problems with restrt's and igb > 0?

For example if I take a simple 2-mer GC-CG basepair DNA and run the
following:

1) 20 steps md from a cold start (irest=0)

2) 10 steps md from a cold start (irest=0) - save restart file. Resume 10
steps md from a hot start (irest=1, ntx=5) from the restart.

If igb=0 then 1 gives exactly the same answer as 2. However, if igb>0,
e.g.
igb=1 or igb=6 (vacuum) then the kinetic energy on step 1 of the restart
in
2 is not the same as the kinetic energy on step 11 of 1.

E.g.

--------- Run 1 -----------
 
--------------------------------------------------------------------------

--
--
 NSTEP =       10   TIME(PS) =       0.020  TEMP(K) =    96.03  PRESS =
0.0
 Etot   =      -378.2090  EKtot   =        34.9209  EPtot      =
-413.1299
 BOND   =       384.7942  ANGLE   =        38.7651  DIHED      =
68.9225
 1-4 NB =        35.0666  1-4 EEL =      -773.9783  VDWAALS    =
-8.6674
 EELEC  =        94.3652  EGB     =      -252.3978  RESTRAINT  =
0.0000
 
--------------------------------------------------------------------------
--
--
 NSTEP =       11   TIME(PS) =       0.021  TEMP(K) =   540.63  PRESS =
0.0
 Etot   =      -401.7712  EKtot   =       196.6025  EPtot      =
-598.3737
 BOND   =       182.9792  ANGLE   =        44.0551  DIHED      =
68.9601
 1-4 NB =        34.7372  1-4 EEL =      -773.5102  VDWAALS    =
-9.2240
 EELEC  =        93.4651  EGB     =      -239.8361  RESTRAINT  =
0.0000
 
--------------------------------------------------------------------------
--
--
 NSTEP =       12   TIME(PS) =       0.022  TEMP(K) =   892.25  PRESS =
0.0
 Etot   =      -419.3046  EKtot   =       324.4722  EPtot      =
-743.7768
 BOND   =        14.2879  ANGLE   =        48.3285  DIHED      =
68.9517
 1-4 NB =        34.3658  1-4 EEL =      -771.6641  VDWAALS    =
-12.1999
 EELEC  =        91.2159  EGB     =      -217.0625  RESTRAINT  =
0.0000
 
--------------------------------------------------------------------------
--
--
---------- Run 2 restart ----------
 NSTEP =        1   TIME(PS) =       0.021  TEMP(K) =   543.24  PRESS =
0.0
 Etot   =      -400.8219  EKtot   =       197.5517  EPtot      =
-598.3737
 BOND   =       182.9792  ANGLE   =        44.0551  DIHED      =
68.9601
 1-4 NB =        34.7372  1-4 EEL =      -773.5102  VDWAALS    =
-9.2240
 EELEC  =        93.4651  EGB     =      -239.8361  RESTRAINT  =
0.0000
 
--------------------------------------------------------------------------
--
--
 NSTEP =        2   TIME(PS) =       0.022  TEMP(K) =   896.25  PRESS =
0.0
 Etot   =      -417.8908  EKtot   =       325.9260  EPtot      =
-743.8168
 BOND   =        14.2351  ANGLE   =        48.3170  DIHED      =
68.9525
 1-4 NB =        34.3470  1-4 EEL =      -771.6661  VDWAALS    =
-12.2822
 EELEC  =        91.3046  EGB     =      -217.0248  RESTRAINT  =
0.0000
 
--------------------------------------------------------------------------
--
--
 NSTEP =        3   TIME(PS) =       0.023  TEMP(K) =   559.41  PRESS =
0.0
 Etot   =      -400.4081  EKtot   =       203.4318  EPtot      =
-603.8399
 BOND   =       138.9595  ANGLE   =        49.7006  DIHED      =
68.8636
 1-4 NB =        33.9926  1-4 EEL =      -769.4735  VDWAALS    =
-16.5652
 EELEC  =        88.7817  EGB     =      -198.0991  RESTRAINT  =
0.0000
 
--------------------------------------------------------------------------
--
--
-----------------
So, everything is the same except the kinetic energy:
 NSTEP =       11   TIME(PS) =       0.021  TEMP(K) =   540.63  PRESS =
0.0
 Etot   =      -401.7712  EKtot   =       196.6025  EPtot      =
-598.3737
 NSTEP =        1   TIME(PS) =       0.021  TEMP(K) =   543.24  PRESS =
0.0
 Etot   =      -400.8219  EKtot   =       197.5517  EPtot      =
-598.3737
Thus after 20 steps you do not reach the same structure. Note, if igb=0
then
the kinetic energies and the final results come out correct. I have tried
this both with Amber 8 and Amber 9.
I have tried this with 3 different systems now and get the same problems
with restarting igb>0 every time.
Does anybody know what is going on here? I can't track it down.
All the best
Ross
/\
\/
|\oss Walker
| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- ross.rosswalker.co.uk |
| http://www.rosswalker.co.uk/ | PGP Key available on request |

Received on Wed Apr 05 2006 - 23:49:59 PDT