Hi Dan,
Sure. I'm using AT1.4 with patches 1-10, compiled with intel 11.1.072, serial version. The system is ~17k atoms, ~3k are protein. I have a total of 6 netCDF trajectories with box info:
1. 38836 frames
2. 6529 frames
3. 2548
4. 8049
5. 7124
6. 53099
The water has not be wrapped at all. Referring to my first email, the frame it crashed on, 56129, occurs on frame 166 of trajectory 5. The trajectory was produced when I striped the water and Na+ from them while concurrently concatenating ie:
trajin 1 through 6
strip wat, na+
trajout all.nc netcdf
Completes no errors (I used proper syntax--not what's above!). I used no other commands in ptraj. This new trajectory has 4 frames on trajectory 5 which get garbled. The energies from mdout and fine and they are fine pre-processing in vmd. I'm not sure what would be the most helpful in terms of which files to give you to try to reproduce. I just reran it and I reproduced it, however the first frame garbled frame occurred at 56128 instead of 56129. The strip command seems to have no affect as I tried it with and without with the same result. I saved the frames as pdbs.. and basically it just looks like a chunk of the coordinates just bottom out to 0 and then goes back to normal (However they are all still present). I suppose GB complained because it was working on a frame with all atoms with 0,0,0 coordinates. There seems to be no pattern because it happens in the middle of a trajectory, starts at the middle of a frame and fixes itself in the middle of a frame.
I'd be more than willing to give you the trajectories to play with but they are ~20gb total. I can try different permutations of them in ptraj and see if it is reproducable with a small number of frames.
Let me know if there is any other information you need.
Trevor
From: Daniel Roe [daniel.r.roe.gmail.com]
Sent: Friday, October 01, 2010 6:15 PM
To: AMBER Mailing List
Subject: Re: [AMBER] sander gb post-processing fail
Hi,
Could you provide some more details so it can be determined if a bug
exists in ptraj or not? When you mean ptraj 'decided to screw with
only 4 frames', does it mean the frames were corrupted, dropped, or
something else? Is it reproducible? What is the size of your system in
atoms and how long are the trajectories you are using? What version of
ptraj are you using, and are all the bugfixes applied? Also, are you
100% certain the original trajectories do not contain any issues?
Sorry for the barrage of questions - I just want to fix any problems
that might be in the code. Thanks!
-Dan
On Fri, Oct 1, 2010 at 3:11 PM, Trevor Gokey <tgokey.mail.sfsu.edu> wrote:
> I just want to follow this up with what happened... for some reason when I combined the TIP3 MD trajectories into one for imin=5 processing, ptraj decided to screw with only 4 frames in the -middle- of one of the trajectories. Ptraj did not report any error when processing and the original trajectories are fine. So, point in hand, if for some reason anyone gets that error message... be sure to check the frame coordinates it failed on.
>
> Trevor
> ________________________________________
> From: Trevor Gokey [tgokey.mail.sfsu.edu]
> Sent: Monday, September 27, 2010 2:25 PM
> To: amber.ambermd.org
> Subject: [AMBER] sander gb post-processing fail
>
> Hello,
>
> I got this line at the bottom (not sure if it's even an error) in my mdout file while using sander.MPI (Amber11) to post process and trajectory.
>
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 1 1.7936E+02 1.7701E+02 5.7539E+03 SG 2696
>
> BOND = 693.8115 ANGLE = 1705.4697 DIHED = 2288.0264
> VDWAALS = 2873.7781 EEL = -15679.1151 EGB = -2811.2453
> 1-4 VDW = 738.9313 1-4 EEL = 10289.4658 RESTRAINT = 0.0000
> ESURF = 80.2363
> minimization completed, ENE= 0.17935881E+03 RMS= 0.177014E+03
> minimizing coord set # 56129
> iatenvcnt .gt. iemax in icosasurf
>
> and it just quit with no other warning as far as I can tell. I've ran a previous trajectory successfully with the same input, which I extracted from mmpbsa.py (except the ioutfm=1 part):
>
> 1ANE 115 ns
> &cntrl
> ntb=0, idecomp=0, cut=999.0, nsnb=99999,
> imin=5, maxcyc=0, ncyc=0, igb=5,
> saltcon=0, intdiel=1, extdiel=78.3,
> gbsa=2, surften=0.0072,ioutfm=1
> &end
>
> What should I do seeing an error like that? The only thing I can extrapolate from the error message is that the (i)nternal(e)nergymax in icosasurf(surften issue) has something wrong. Am I far off?
>
> Thanks,
> Trevor Gokey
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Oct 01 2010 - 21:30:03 PDT
