AMBER Archive (2009)

Subject: Re: [AMBER] errors on REMD

From: z g (zgong.hust_at_gmail.com)
Date: Thu Jan 22 2009 - 19:02:08 CST

Dear sir :
       I have followed your advise to change my mdin files like this
Equilibration
 &cntrl
   irest=1, ntx=5,
   nstlim=10000, dt=0.001,
   ntt=3, gamma_ln=1.0,
   temp0=XXXXX, ig=RANDOM_NUMBER,
   ntc=2, ntf=2, nscm=1000,
   ntb=0, igb=1,
   cut=999.0, rgbmax=999.0,
   ntpr=1000, ntwx=1000, ntwr=100000,
   nmropt=1,
   numexchg=1000,
 /
 &wt TYPE='END'
 /
However, the same erros happened again

input conversion error, unit 9, file /export/home/md/remd.rst.007

the remd.rst.007 file like

  389 0.1360000E+04 0.3756000E+03
 -67.9667402 143.5693955 135.3920753 26.3374926 259.7714224-374.0863183
  27.2250652 258.8290660-374.5986003-138.7437384-753.0304527 -24.6733927
 578.0399447 544.7979446-203.7098287 27.3056520 258.7732931-376.0732231
-673.6576239 334.1924531-697.6538998 28.2680961 259.7702090-376.4543691
  29.0208772 259.2781126-377.5736472 42.6589530 585.2045901-745.6005556
  30.4860660 258.9320589-377.4628697 31.2380916 258.3566393-378.4070212
-302.2943711 -56.2521559************ 32.5331551 258.2713939-378.2040061

before I do the remd, I have I have run 1ns simulations with each replica to
equilibrate them to their individual temperatures
my amber version is amber9
what should I do ?

Thank you very much!

2009/1/20 Carlos Simmerling <carlos.simmerling_at_gmail.com>

> I don't think your main problem is unique to remd. however, your
> inputs will probably not work properly for remd. make sure to read the
> manual carefully, especially the section "cautions when using replica
> exchange".
>
> one of your problems is specifically discussed there:
> irest should be 1, not 0. ntx should be 5. you almost certainly don't
> want to reassign velocities each exchange. this will very likely
> interfere with correct equilibration and probably cause exchanges to
> be incorrect.
> Amber10 has much improved input checks and other improvements for remd.
>
> regarding your problem with coordinates, this is not REMD related but
> indicates that your system has diffused so far from the origin that
> the coordinates don't fit in the formatted file. I have never had this
> happen with ntt=3, though I guess it is possible. using nscm>0 will
> resolve it, though should be needed for ntt=3. did you run with ntt=0
> or 1 at some point? is there a reason the system would be so far from
> the origin?
>
>
>
> On Tue, Jan 20, 2009 at 8:39 AM, z g <zgong.hust_at_gmail.com> wrote:
> > I'm using the "amber9" for REMD
> > Thank you very much!
> >
> > 2009/1/20 Carlos Simmerling <carlos.simmerling_at_gmail.com>
> >
> >> your input looks like it may have some problems, but I cannot know
> >> unless you tell me which amber version you are using.
> >>
> >> On Mon, Jan 19, 2009 at 9:41 PM, z g <zgong.hust_at_gmail.com> wrote:
> >> > Dear sir :
> >> > I would like to use amber9 to perform REMD on RNA molecule.
> >> However,
> >> > I met some errors during the simulation.
> >> > At first, Before starting the REMD simulations we have run 1ns
> >> simulations
> >> > with each replica to equilibrate them to their individual
> temperatures. I
> >> > have use 8 replicas with temperatures(
> 270,285,300,315,330,345,370,380).
> >> The
> >> > RNA molecule contains 389 atoms.
> >> > My mdin files for REMD like
> >> >
> >> > RNA REMD
> >> > &cntrl
> >> > irest=0, ntx=1,
> >> > nstlim=10000, dt=0.001,
> >> > irest=0, ntt=3, gamma_ln=1.0,
> >> > temp0=287.3, ig=7988,
> >> > ntc=2, ntf=2,
> >> > ntb=0, igb=1,
> >> > cut=999.0, rgbmax=999.0,
> >> > ntpr=1000, ntwx=1000, ntwr=100000,
> >> > nmropt=1,
> >> > numexchg=1000,
> >> > /
> >> > &wt TYPE='END'
> >> > /
> >> > At first, the program run correctly, while after 1 hour, some errors
> >> happen
> >> >
> >> > forrtl: severe (64): input conversion error, unit 9, file
> >> > /export/home/1zih_remd//remd.rst.007
> >> >
> >> > I have watched the remd.rst.007, and find the numbers are very large
> >> >
> >> >
> >> > 389 0.1300000E+04 0.3592000E+03
> >> > -105.8953861 748.7806467-468.5337764-100.7161016
> 24.3022386-170.1731608
> >> > -101.4099610 24.3477791-168.9152646-471.2815199
> -14.2248508-271.6494052
> >> > -49.2667843 548.1655243 82.0420727-101.1563083
> 23.2039701-167.8895453
> >> > -213.6132285 301.0739486 451.6375036-102.1736480
> 22.2559916-168.1535552
> >> > -101.4738436 20.9869045-168.1231668 -75.7654453-315.9162929
> 590.7650174
> >> > -102.0262742 20.0482355-169.1328678-101.5804244
> 19.7278654-170.3871519
> >> > 534.9756401 508.7310783-616.9927623-102.2027898
> 18.7924921-171.0093015
> >> > -103.0636418 18.3286643-169.9952243-104.0138694
> 17.1960597-170.0369936
> >> > -104.3907387 16.3896575-170.8753573-104.6018766
> 16.9349549-168.7753226
> >> > -503.6963882-583.9394758 -15.3460633-104.5414249
> 17.8285116-167.7480043
> >> > -105.1545082 17.4537929-166.6067724-535.9608133-746.2963594
> -95.6632087
> >> > 171.1120712 444.0063185 577.8922425-103.6933133
> 18.8393584-167.6861132
> >> > -103.0122636 19.0754130-168.8968330 -99.7914615
> 22.5807460-167.9552119
> >> > 417.3237163 541.2504776-584.2170254-100.0500924
> 21.2862850-168.6877856
> >> > -252.8170614 80.7117256************ -99.0037521
> 20.3504289-168.5717094
> >> > 443.9874455 196.9228391-437.3244152 -99.3986068
> 22.4688989-166.6403937
> >> >
> >> > What should I do to fix this problem?
> >> > Thank you !
> >> > _______________________________________________
> >> > AMBER mailing list
> >> > AMBER_at_ambermd.org
> >> > http://lists.ambermd.org/mailman/listinfo/amber
> >> >
> >>
> >> _______________________________________________
> >> AMBER mailing list
> >> AMBER_at_ambermd.org
> >> http://lists.ambermd.org/mailman/listinfo/amber
> >>
> > _______________________________________________
> > AMBER mailing list
> > AMBER_at_ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber