Re: [AMBER] Unit cell size changed and coords overlap problems

From: Daniel Roe <daniel.r.roe.gmail.com>
Date: Fri, 11 Mar 2022 11:44:10 -0500

Hi, I'm not sure you saw my earlier response to this so I'll post it again.

As the messages indicate, certain frames in your trajectory file are
corrupted (with NetCDF files this usually manifests as all coordinates
being 0, which cpptraj is catching here). Those frames are not usable
and should be discarded.

>
> terminate called after throwing an instance of 'std::bad_alloc'

I think this was a bug in the 'check' command that I fixed. What
version of cpptraj are you using. If not the GitHub version, can you
try using that and see if the behavior is fixed? Thanks,

-Dan

On Thu, Mar 10, 2022 at 9:01 PM <chanyann06.stu.scu.edu.cn> wrote:
>
> Dear Amber Users,
>
> I carried out 2us GaMD of protein, which is composed of ten 200ns-MD trajectories. However, in the second trajectory, when I calculated distance between two residues, I got warnings as followed:
>
> ----- GaMDprod_02.nc (1-40000, 1) -----
>
> 0% 10% 20% 30% Warning: Frame 14529 imaging failed, box lengths are zero.
>
> Warning: Frame 14530 coords 1 & 2 overlap at origin; may be corrupt.
>
> Warning: Frame 14530 imaging failed, box lengths are zero.
>
> ……
>
> ……
>
> ……
>
> Warning: Frame 25099 coords 1 & 2 overlap at origin; may be corrupt.
>
> Warning: Frame 25099 imaging failed, box lengths are zero.
>
> 70% 80% 90% 100% Complete.
>
>
>
>
>
>
>
> Here is my .in file used to run the simulation:
>
> GaMD production
> &cntrl
> imin = 0,
> irest = 0,
> ntx = 1,
> ioutfm = 1,
> nstlim = 100000000,
> dt = 0.002,
> ntt = 3,
> gamma_ln = 5.0,
> ig = -1,
> tempi = 310.0,
> temp0 = 310.0,
> ntp = 0,
> ntb = 1,
> ntc = 2,
> ntf = 2,
> cut = 12,
> ntwr = 2500,
> ntpr = 2500,
> ntwx = 2500,
> ntwe = 2500,
> iwrap = 1,
> ntr = 0,
> ntwprt = 0,
>
> igamd = 3, iE = 1, irest_gamd = 1,
> ntcmd = 0, nteb = 0, ntave = 600000,
> ntcmdprep = 0, ntebprep = 0,
> sigma0P = 6.0, sigma0D = 6.0,
> &end
> /
>
>
>
> Here is the .in file I used to calculate the distance:
>
> parm step5_input.parm7
>
> trajin /home/xmpu233/cx/DAM128paraneo/GaMDprod_02.nc 1 last
>
> center origin :296-583
> image origin center
> strip :POPC
> strip :CHL1
> strip :SOD
> strip :CLA
> strip :WAT
> reference /home/xmpu233/cx/DAM128paraneo/step5_input.rst7
> distance :396.CA,C,O,N,H :510.CA,C,O,N,H out dis2.dat
>
>
> I have extracted a warning frame and found that all atom coordinates changes to zero, like this:
>
> CRYST1 0.000 0.000 0.000 0.00 0.00 0.00 1
> ATOM 1 N ARG 1 0.000 0.000 0.000 1.00 0.00 N
> ATOM 2 HT1 ARG 1 0.000 0.000 0.000 1.00 0.00 H
> ATOM 3 HT2 ARG 1 0.000 0.000 0.000 1.00 0.00 H
> ATOM 4 HT3 ARG 1 0.000 0.000 0.000 1.00 0.00 H
> ATOM 5 CA ARG 1 0.000 0.000 0.000 1.00 0.00 C
> ATOM 6 HA ARG 1 0.000 0.000 0.000 1.00 0.00 H
> ATOM 7 CB ARG 1 0.000 0.000 0.000 1.00 0.00 C
> ATOM 8 HB1 ARG 1 0.000 0.000 0.000 1.00 0.00 H
> ATOM 9 HB2 ARG 1 0.000 0.000 0.000 1.00 0.00 H
> ATOM 10 CG ARG 1 0.000 0.000 0.000 1.00 0.00 C
> ATOM 11 HG1 ARG 1 0.000 0.000 0.000 1.00 0.00 H
> ATOM 12 HG2 ARG 1 0.000 0.000 0.000 1.00 0.00 H
>
>
>
>
> "check" has been used in cpptraj to check the frame and I got this message:
>
> Warning: Unit cell size has changed so much that grid must be recalculated.
>
> Warning: Old sizes= {84, 84, 77} New sizes= {1, 1, 1}
>
> Warning: X cell offset reset to 1
>
> Warning: Y cell offset reset to 1
>
> Warning: Z cell offset reset to 1
>
> Total Grid memory: 0.208 kB
>
> terminate called after throwing an instance of 'std::bad_alloc'
>
> what(): std::bad_alloc
>
>
>
>
>
>
> It is confusing that the frames before and after the warning frames (14530-25099 frames report warnings) seem normal. Can the frame 14529 ( the one before warning) and frame 25100 (the one after warning) be continuous? What would cause this warning?
>
> I've check similar problems in amber mailing list but could not sovle it.
>
> Looking forward to your help and thanks a lot!
>
>
>
>
>
>
> _______________________________________________
> 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 Mar 11 2022 - 09:00:02 PST
Custom Search