Re: [AMBER] Problem with simulation of protein-DNA complex

From: Aditya Sarkar <aditya.molbio.gmail.com>
Date: Mon, 15 Dec 2014 16:40:59 +0530

Hi Jason,

Thanks a lot for your reply. I have go through your's and Daniel's
suggestion and found followings:

*NOTE: I have used distance restraint on the DNA to prevent fraying of DNA
ends.*

*DNA force field issue*: I have found that the recent DNA related force
field is XOL4 that includes bsc0 modification along with the improvements
related to Quadruplex DNA (please correct me if I wrong). I have also found
in literature that the B-DNA is quite stable (except the ends) in bsc0
force field (Ref:Base Pair Fraying in Molecular Dynamics Simulations of DNA
and RNA, J. Chem. Theory Comput., 2014, 10 (8), pp 3177–3189)

*Is this a imaging problem:* As you say it may be a imaging problem, but I
do image the entire trajectory using "autoimage" and then visualized it. I
have also calculate the DNA RMSD (P atom) from the EM structure and is
given below:

RMSD of DNA (P atom) from autoimaged trajectory of *prodrun47.mdcrd*
and *prodrun48.mdcrd
(the DNA takes horrible structure in this mdcrd file)*:

 #Frame RMSD_00002
       1 3.4768
       2 3.3383
       3 3.4334
       4 3.3140
       5 3.3396
       6 3.1830
       7 3.1468
       8 3.1713
       9 3.0603
      10 3.0466
      //
     388 2.6668
     389 2.7159
     390 2.7442
     391 2.7269
     392 2.6591
     393 2.7106
     394 2.7090
     395 2.6590
     396 2.7572
     397 2.7799
     398 2.9221
     399 3.0322
     400 2.9314................*.end of prodrun47.mdcrd*

     401 11.3305................*starting of prodrun48.mdcrd*
     402 17.0298
     403 20.1587
     404 21.2551
     405 21.6177
     406 21.7009
     407 21.8027
     408 21.8476
     409 22.0402
     410 22.0547
    //


*Does the DNA get denatured:* I am also in doubt in this case because of
two reasons:

1. If you kindly consider the above data, in the very beginning of
prorun48.mdcrd (#Frame 401) the RMSD value get suddenly high. I have save
each snapshot in 5ps interval and within that interval is it possible that
5-6 bases from each end of DNA get denatured (my DNA is 36 residue long) ???

2. *When I redo the simulation by imaging the prodrun47.rs
<http://prodrun47.rs>*t (as suggested by Dan and you) and visualize the
trajectory it was fine. The RMSD is as follow:

RMSD of DNA (P atom) from autoimaged trajectory of prodrun47.mdcrd and
prodrun48.mdcrd (the DNA takes horrible structure in this mdcrd fiel):

 #Frame RMSD_00002

     388 2.6668
     389 2.7159
     390 2.7442
     391 2.7269
     392 2.6591
     393 2.7106
     394 2.7090
     395 2.6590
     396 2.7572
     397 2.7799
     398 2.9221
     399 3.0322
     400 2.9314 ...............end of prodrun47.mdcrd

     401 2.9772................starting of prodrun48.mdcrd
     402 2.8694
     403 2.9449
     404 3.0418
     405 2.9887
     406 2.8980
     407 2.9807
     408 2.7438
     409 2.9992
     410 2.8065

*One interesting Observation:*

*Without doing autoimage* if I calculate RMSD from my initial
prodrun47.mdcrd and prodrun48.mdcrd by using only "RMS reference" option *I
have found an interesting observation at the end of the prodrun47.mdcrd (at
the 400th frame): the RMSD is high*

 #Frame RMSD_00002
     388 2.6668
     389 2.7159
     390 2.7442
     391 2.7269
     392 2.6591
     393 2.7106
     394 2.7090
     395 2.6590
     396 2.7572
     397 2.7799
     398 2.9221
     399 3.0322
     *400 49.3009* .....................end of prodrun47.mdcrd

     401 45.3942................starting of prodrun48.mdcrd
     402 42.5245
     403 40.0904
     404 38.4978
     405 38.0837
     406 37.9209
     407 37.7905
     408 37.6711
     409 37.3429
     410 37.3395

*What I understood*: This might be a imaging issue of restart file.

But I do not understood why this happens????????????

Please help me if possible.


Regards
Aditya







On Sat, Dec 13, 2014 at 9:20 PM, Indrajit Deb <biky2004indra.gmail.com>
wrote:
>
>
>
>
>
>
> ---------------------------------------------------------------------
> Indrajit Deb
> Kolkata, India.
> Mob: +919239202278
>
> ---------- Forwarded message ----------
> From: Jason Swails <jason.swails.gmail.com>
> Date: Sat, Dec 13, 2014 at 8:40 PM
> Subject: Re: [AMBER] Problem with simulation of protein-DNA complex
> To: AMBER Mailing List <amber.ambermd.org>
> Cc: daniel.r.roe.gmail.com
>
>
> > On Dec 13, 2014, at 1:12 AM, Aditya Sarkar <aditya.molbio.gmail.com>
> wrote:
> >
> > Hi Dan,
> >
> > Thanks for your answer. I forgot to mentioned that I have done "autoimge"
> > on the entire trajectory and that horrible structure appeared suddenly.
> >
> > But I have few problems again:
> >
> > as you suggest to do re-imging the restart file using "autoimage", but I
> > can not figure out how to do that. I read the manual and planned to do
> > using
> >
> > trajin prodrun47.rst
> > autoimage
> > trajout imagedprodrun47.rst
> >
> > *Problem1*: Does that restart file retain the time and velocity
> > information?
>
> With the latest version of cpptraj, yes.
>
> > I found that the writing format is different in the newly generate
> restart
> > file and is not recognized by pmemd.
>
> You need to either change the imagedprodrun47.rst filename to an extension
> recognized by cpptraj (e.g., .rst7) or specify the output file format on
> the trajout line. e.g.:
>
> trajout imagedprodrun47.rst restart
>
> > Should I use "remdtraj" option? but I am unable to figure out the correct
> > syntax?
>
> Are you using REMD? If not, then no.
>
> > *Problem2: *If this problem solved by re-imaging the restart file, do I
> > need to redo the entire simulation from beginning by re-imaging each
> > restart file or I can finish the last few runs of this simulation by
> doing
> > re-imaging?
>
> No. Imaging just affects visualization, not the actual results. The only
> effect it can have on the simulation is if you are using restraints, which
> do not respect periodic images. If you are hit with this issue, though,
> your simulation will usually explode almost immediately (since the
> restraint distance jumps by the length of the box, resulting in enormous
> forces).
>
> > *Problem3:*Two other MD simulations of this system in mutated form were
> > done. They do not show such anomaly. Do I need to do the same re-imaging
> > for them also?
>
> Only if your analyses require that.
>
> It sounds to me like your DNA denaturation is real. You should make sure
> that the force field you’re using (ff99SBildn+bsc0) is really the “best”
> and latest one. ff99SBildn is a protein force field -- it is equivalent to
> ff99SB+bsc0 for DNA. See if the Amber manual has some information about
> whether some of the newer force fields (like ff10 or ff14SB) contain more
> updated DNA parameters (I know their RNA parameters are updated, but I’m
> not sure about DNA).
>
> At this point, you may have to look for a scientific explanation about
> your observations (rather than a possible technical explanation describing
> a visualization artifact).
>
> HTH,
> Jason
>
>
> >
> > Looking forward for your reply.
> >
> > regards
> > Aditya
> >
> >
> >
> >
> > On Fri, Dec 12, 2014 at 8:25 PM, Indrajit Deb <biky2004indra.gmail.com>
> > wrote:
> >>
> >>
> >>
> >>
> >>
> >>
> >> ---------------------------------------------------------------------
> >> Indrajit Deb
> >> Kolkata, India.
> >> Mob: +919239202278
> >>
> >> ---------- Forwarded message ----------
> >> From: Daniel Roe <daniel.r.roe.gmail.com>
> >> Date: Fri, Dec 12, 2014 at 8:20 PM
> >> Subject: Re: [AMBER] Problem with simulation of protein-DNA complex
> >> To: AMBER Mailing List <amber.ambermd.org>
> >>
> >> Hi,
> >>
> >> Have you looked at your entire trajectory? This sounds like it could
> >> be an imaging artifact to me. Because the DNA is two separate
> >> molecules, occasionally one strand gets imaged and the other doesn't.
> >> This has no affect during the simulation since it's just for
> >> visualization, but it is particularly bad if you have restraints that
> >> depend on the DNA molecules because restraints are *not* imaged.
> >> Whenever I simulate multiple solute molecules with PBC and restraints
> >> I always re-image my coordinate files before restarting. You should
> >> try re-imaging your restart with cpptraj's 'autoimage' command (the
> >> one you used as coordinates for the run that blew up) and see if that
> >> helps.
> >>
> >> If on the other hand you've viewed your trajectory and the DNA is
> >> actually falling apart gradually (i.e. there is no sudden "jump" from
> >> imaging) that's another story.
> >>
> >> -Dan
> >>
> >> On Fri, Dec 12, 2014 at 1:49 AM, Aditya Sarkar <aditya.molbio.gmail.com
> >
> >> wrote:
> >>> Dear Amber users,
> >>>
> >>> I am new to the MD simulation. I have faced a strange problem in doing
> >>> simulation of protein-DNA complex. The problem, questions and the setup
> >>> details are given below:
> >>>
> >>> *****PROBLEM********
> >>> ---------------------------------
> >>>
> >>> I have planned to run 100ns simulation and divide the whole run in 50
> >>> restart (each of 2ns). My simulation ran well upto 94ns but at the
> >> begining
> >>> of the next restart I found the DNA get horribly denatured. I have
> >> attached
> >>> the picture with this mail.
> >>>
> >>> *****QUESTION********
> >>> ---------------------------------
> >>>
> >>> 1. I have checked the prmtop file with rdparm to make sure about the
> use
> >> of
> >>> proper force field. But I have doubt whether my way of selection of
> force
> >>> field is correct or not? and also regarding the production run script.
> >>>
> >>> 2. I did not get why the DNA get unstable and how can I analyze the
> cause
> >>> and correct it?
> >>>
> >>>
> >>> *****SETUP********
> >>> ---------------------------
> >>>
> >>> 1. The structure is a crystal structure with all crystal water. DNA
> >> residue
> >>> no: 103-138
> >>>
> >>> *2.Force field used: ff99SB-ILDN along with parambsc0*
> >>>
> >>> *[tleap file command: source leaprc.ff99SBildn*
> >>> * loadoff DNA_CI.lib*
> >>> * loadamberparams frcmod.parmbsc0*
> >>>
> >>>
> >>> *I do did energy minimization (with and without position restraint,
> >> heating
> >>> (with position restraint), NPT equilibration (gradually relaxing
> >> restraint)
> >>> and finally performed production run*
> >>>
> >>> *3.Production run:*
> >>> &cntrl
> >>> imin = 0, irest = 1, ntx = 5,
> >>> ntb = 2, pres0 = 1.0, ntp = 1,
> >>> taup = 2.0,
> >>> cut = 8.0, ntr = 0,
> >>> ntc = 2, ntf = 2,
> >>> tempi = 300.0, temp0 = 300.0,
> >>> ntt = 3, gamma_ln = 1.0, ig = -1, iwrap = 1,
> >>> nstlim = 1000000, dt = 0.002, nmropt =1,
> >>> ntpr = 2500, ntwx = 2500, ntwr = 2500,
> >>> ntwv = 2500, ntwe = 2500
> >>> /
> >>> &wt type='DUMPFREQ', istep1=2000
> >>> /
> >>> &wt type='END'
> >>> /
> >>> LISTOUT=POUT
> >>> DISANG=DIST.rst
> >>> DUMPAVE=rest_values_INPRORUN1
> >>> END
> >>>
> >>> *4. Distance restraint input file*
> >>> 2 ALA O 6 ALA N 3.0
> >>> 3 LEU O 7 ARG N 3.0
> >>> 53 ALA O 57 ALA N 3.0
> >>> 54 LEU O 58 ARG N 3.0
> >>> 50 LEU N 46 ARG O 2.9
> >>> 49 LYS N 45 ALA O 3.2
> >>> 101 LEU N 97 ARG O 3.0
> >>> 100 LYS N 96 ALA O 3.0
> >>> 103 DG5 O6 138 DC3 H41 2.0
> >>> 103 DG5 H1 138 DC3 N3 1.9
> >>> 103 DG5 H22 138 DC3 O2 3.5
> >>> 121 DG5 O6 120 DC3 H41 1.9
> >>> 121 DG5 H1 120 DC3 N3 1.9
> >>> 121 DG5 H22 120 DC3 O2 3.5
> >>>
> >>> *5. Commands used:*
> >>>
> >>> to generate DIST.rst file:
> >>> ****************************
> >>> makeDIST_RST -upb DNA_Protein-endDISTRST -pdb emin_NoRST.pdb -rst
> >> DIST.rst
> >>>
> >>>
> >>> For production run:
> >>> **********************
> >>> nohup mpiexec.hydra -n 8 sander.MPI -O -i production1.in -o
> >> prodrun1.out -p
> >>> GCN4_CREB_inbox.prmtop -c equil_NPT.rst -r prodrun1.rst -x
> prodrun1.mdcrd
> >>>
> >>> nohup mpiexec.hydra -n 8 pmemd.MPI -O -i production2.in -o
> prodrun2.out
> >> -p
> >>> GCN4_CREB_inbox.prmtop -c prodrun1.rst -r prodrun2.rst -x
> prodrun2.mdcrd
> >>>
> >>> *NOTE: I was initially unsure about using "pmemd" that's why I started
> >>> first run with sander and after reading maual and Amber blog, I started
> >>> using "pmemd".
> >>>
> >>>
> >>> Looking forward for your reply.
> >>>
> >>> regards
> >>> Aditya
> >>>
> >>> _______________________________________________
> >>> AMBER mailing list
> >>> AMBER.ambermd.org
> >>> http://lists.ambermd.org/mailman/listinfo/amber
> >>>
> >>
> >>
> >>
> >> --
> >> -------------------------
> >> Daniel R. Roe, PhD
> >> Department of Medicinal Chemistry
> >> University of Utah
> >> 30 South 2000 East, Room 307
> >> Salt Lake City, UT 84112-5820
> >> http://home.chpc.utah.edu/~cheatham/
> >> (801) 587-9652
> >> (801) 585-6208 (Fax)
> >>
> >> _______________________________________________
> >> 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
>
> --
> Jason M. Swails
> BioMaPS,
> Rutgers University
> Postdoctoral Researcher
>
> _______________________________________________
> 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 Mon Dec 15 2014 - 03:30:04 PST
Custom Search