This is to be expected, as the restart file is 1 step ahead of the
final mdcrd I believe. You can check this by rerunning exactly the
same calculation for N steps, and compare that to the N+1th frame of
the mdcrd from your first simulation. It may not match exactly due to
the fact that the precision of a restart is different than a mdcrd,
but it should match to the precision of the mdcrd.
Hope this helps,
Jason
On Mon, Dec 14, 2009 at 8:23 PM, Sidney Elmer <paulymer.gmail.com> wrote:
> Hi,
>
> I am getting inconsistencies when comparing the coordinates from the restart
> file generated by sander and the output files generated from ptraj. See
> below for a summary of what I did. Is this a bug or am I missing
> something? Thanks for any insights.
>
> Sid
>
> 1. setup a run with 500 steps; restart file updated every 100 steps:
> $ cat mdin
> &cntrl
> nstlim = 500, dt = 0.002,
> ntx = 5, irest = 1,
> ntf = 2, ntc = 2,
> ntt = 3, temp0 = 310.0, gamma_ln = 5.0,
> ntb = 2, ntp = 1, pres0 = 1.0,
> ntwx = 1, ntwr = 100, ntpr = 1,
> /
>
> 2. execute run:
> $ sander -O
>
> 3. thus, after execution finishes, the coordinates in "restrt" file should
> match coordinates from frame 500, correct? To test this, I dumped the
> contents of "mdcrd" into separate files for each frame:
> $ cat ptraj.cmd
> trajin mdcrd
> trajout out.restrt restart
>
> $ ptraj prmtop < ptraj.cmd
>
> 4. compare coordinates from "restrt" file with the coordinates from frame
> 500:
> $ head restrt out.restrt.500
> ==> restrt <==
>
>
> 12229
> 36.3305038 13.8660564 76.3226335 36.8364002 13.4737666 77.1038361
> 35.6825593 13.1233185 76.1021414 35.8160375 14.7061467 76.5455088
> 37.2998211 13.9844329 75.2107251 37.9789653 13.1328655 75.2519122
> 36.6052142 13.9179318 73.8372047 35.9462419 13.0519369 73.7746877
> 36.0497970 14.8077383 73.5407853 37.4626397 13.7198942 72.5764882
> 38.1500466 12.8983887 72.7782423 36.8921438 13.5181536 71.6698818
> 38.6032029 15.1222338 72.0682800 38.9906380 14.4044642 70.4856256
> 38.0884652 14.4689332 69.8773264 39.6575401 15.1439896 70.0424002
>
> ==> out.restrt.500 <==
> trajectory generated by ptraj
> 12229
> 36.3300018 13.8699999 76.3199997 36.8400002 13.4700003 77.0999985
> 35.6800003 13.1199999 76.0999985 35.8199997 14.7100000 76.5500031
> 37.2999992 13.9799995 75.2099991 37.9799995 13.1300001 75.2500000
> 36.6100006 13.9200001 73.8399963 35.9500008 13.0500002 73.7699966
> 36.0499992 14.8100004 73.5400009 37.4599991 13.7200003 72.5800018
> 38.1500015 12.8999996 72.7799988 36.8899994 13.5200005 71.6699982
> 38.5999985 15.1199999 72.0699997 38.9900017 14.3999996 70.4899979
> 38.0900002 14.4700003 69.8799973 39.6599998 15.1400003 70.0400009
>
> 5. I then wrote a script that calculated the difference between each element
> in the coordinate vectors to see if there was a systematic change in the
> coordinates; the differences are random distributed between -0.005 and
> +0.005, which are not negligible differences.
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
---------------------------------------
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Dec 14 2009 - 18:00:02 PST
