AMBER Archive (2005)

Subject: Re: AMBER: Floating Exception in sander

From: Robert Duke (rduke_at_email.unc.edu)
Date: Fri Jan 07 2005 - 07:15:44 CST


Anshul -
I have not replied since our last exchange for several reasons: 1) the fpe
problem is not interesting until all the other stuff is cleared up; it could
be due to an incorrect prmtop file generated by leap, 2) the info provided
is insufficient to judge why model building works on one system and not on
the other, but the fact that this is the case indicates there is a problem
with the amber installation on the alpha, or that the procedures used were
not truly identical. I don't have the model, the system that works, or the
one that doesn't. You need to step through the procedure of getting the
model constructed, and see where the first difference occurs. 3) If an fpe
problem remains, more info will be needed to debug it (like the location in
the code where the fpe occurred, and the data that caused the problem). It
is entirely possible that this all relates to real build problems on the
alpha, but if you get through all the test cases under the "test" directory,
that is a preliminary, but not definitive indication that everything is okay
(ie., there are many scenarios not covered by the test cases, and things can
happen in longer runs and at higher processor count that are not covered by
the test cases). Now, I do have more problems getting things to run on the
alpha's than anything else, so the compiler + chip seems in some ways
touchy. Also, I don't typically build the model building s/w (leap et al.)
on every machine, just on an sgi or linux single processor workstation I
have ready access to. SO if you want to get to the bottom of this, you are
going to have to produce some very specific information about what is
failing (ie., first I executed these commands on both of these systems, and
the output differs thusly - <data>). The data below is the same stuff we
got before; it may be a page of info, but it is nowhere near enough to
really solve the problem. I hope someone else who routinely builds
nonstandard models on the list will be able to jump on this when you reply;
that is not my usual activity.
Good luck - Bob

----- Original Message -----
From: <anshul_at_imtech.res.in>
To: <amber_at_scripps.edu>
Sent: Friday, January 07, 2005 9:14 AM
Subject: Re: AMBER: Floating Exception in sander

> Bob Duke and David,
> Thanks for your reply. Actually I was running the job on the same molecule
> previously on SUN system and now I wanted to shift it to this UNIX
> machine. but the behaviour is totally different. My molecule consists of a
> protein along with an organic molecule. I had prepared the frcmod and
> prepin files for ths=is organic molecule. Now when I load the molecule
> after loading the prepin and frcmod files, it complains about some bond
> and angle parameters also says that it can not read the entered parameters
> for the dihedral angles. So I added the missing angle and bond parameters
> and also removed the entered dihedral parameters. now it was able to load
> the molecule and save the parameter and coordinate files. I dont know why
> there is this difference. The frcmod file used by me previopusly is same
> for working on all other machines (SUN, SGI-orign, linux).
> Now coming to the problem of floating exception. the input files used is
> as follows:
> --------------------------------------------
> &cntrl
> imin=1,
> maxcyc=6000, ncyc=4000, drms=0.001,
> ntb=0, ntr=1, ipol=0,
> cut=999, igb=1
> ntpr=500, lastist=3000000,
> &end
> Group input for restrained atoms 15
> RES 1 112
> END
> Group input for restrained atoms
> 15
> RES 116 125
> END
> END
> ------------------------------------------------
>
> The output which I got in responce to it is as follows:
>
>
> -------------------------------------------------------
> Amber 7 SANDER Scripps/UCSF 2002
> -------------------------------------------------------
> Wed Jan 5 21:49:11 2005
> File Assignments:
> MDIN: bsa-nocut-gbmin1.in
> MDOUT: bsa-nocut-gbmin1.out
> |INPCRD: mod2.inpcrd
> | PARM: mod2.prmtop
> |RESTRT: bsa-nocut-gbmin1.rst
> | REFC: mod2.inpcrd
>
> Here is the input file:
> &cntrl
> imin=1,
> maxcyc=6000, ncyc=4000, drms=0.001,
> ntb=0, ntr=1, ipol=0,
> cut=999, igb=1
> ntpr=500, lastist=3000000,
> &end
> Group input for restrained atoms 15
> RES 1 112
> END
> Group input for restrained atoms
> 15
> RES 116 125
> END
> END
> ---------------------------------------------------------------------------
> 1. RESOURCE USE:
> ---------------------------------------------------------------------------Flags:
> |
> Warning: lastist only works when MEM_ALLOC is set
> | at compile time. Resetting the size of the
> | stack to the value from sizes.h: 2000000
> | New format PARM file being parsed.
> | Version = 1.000 Date = 01/03/05 Time = 22:26:52
> NATOM = 9293 NTYPES = 15 NBONH = 4596 MBONA = 4774
> NTHETH = 10574 MTHETA = 6442 NPHIH = 19691 MPHIA = 12015
> NHPARM = 0 NPARM = 0 NNB = 51428 NRES = 602
> NBONA = 4774 NTHETA = 6442 NPHIA = 12015 NUMBND = 53
> NUMANG = 111 NPTRA = 43 NATYP = 40 NPHB = 0
> IFBOX = 0 NMXRS = 37 IFCAP = 0 NEXTRA = 0
> | Memory Use Allocated Used
> | Real 2000000 381385
> | Hollerith 400000 56362
> | Integer 3200000 452707
> | Duplicated 0 dihedrals
> | Duplicated 0 dihedrals
>
> ---------------------------------------------------------------------------
> 2. CONTROL DATA FOR THE RUN
> ---------------------------------------------------------------------------General
> flags:
> imin = 1, nmropt = 0
>
> Nature and format of input:
> ntx = 1, irest = 0, ntrx = 1
> Nature and format of output:
> ntxo = 1, ntpr = 500, ntrx = 1, ntwr =
> 500
> iwrap = 0, ntwx = 0, ntwv = 0, ntwe =
> 0
> ioutfm = 0, ntwprt = 0, idecomp = 0, rbornstat=
> 0
> Potential function:
> ntf = 1, ntb = 0, igb = 1, nsnb =
> 25
> ipol = 0, gbsa = 0
> dielc = 1.00000, cut = 999.00000, intdiel = 1.00000
> saltcon = 0.00000, offset = 0.09000, gbalpha= 1.00000
> gbbeta = 0.00000, gbgamma = 0.00000, surften = 0.00500
> scnb = 2.00000, scee = 1.20000
> Frozen or restrained atoms:
> ibelly = 0, ntr = 1
> Energy minimization:
> maxcyc = 6000, ncyc = 4000, ntmin = 1
> dx0 = 0.01000, dxm = 0.50000, drms = 0.00100
> LOADING THE CONSTRAINED ATOMS AS GROUPS
> 5. REFERENCE ATOM COORDINATES
> ----- READING GROUP 1; TITLE:
> Group input for restrained atoms (1-112)
>
> GROUP 1 HAS HARMONIC CONSTRAINTS 15.00000
> GRP 1 RES 1 TO 112
> Number of atoms in this group = 1729
> ----- READING GROUP 2; TITLE:
> Group input for restrained atoms (116-125)
> GROUP 2 HAS HARMONIC CONSTRAINTS 15.00000
> GRP 2 RES 116 TO 125
> Number of atoms in this group = 144
> ----- END OF GROUP READ -----
>
> ---------------------------------------------------------------------------
> 3. ATOMIC COORDINATES AND VELOCITIES
> ---------------------------------------------------------------------------
> begin time read from input coords = 0.000 ps
>
> Number of triangulated 3-point waters found: 0
>
> ---------------------------------------------------------------------------
> 4. RESULTS
> ---------------------------------------------------------------------------
>
>
> The machine I am using is a digital unix with single processor and alpha
> CPU. the OS version is 4.0 D. the amber version I am using is Amber-7 and
> the compiler used GCC (g77). I even checked the installation by running
> the test case but it passed all the tests.
> I hope I am able to give all the details and address the problem well.
> Thanks once again for your help.
>
> Waiting to hear from you.
> With best regards,
> Anshul
>
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu