AMBER Archive (2002)

Subject: igb problem in sander

From: Emma.Sigfridsson_at_astrazeneca.com
Date: Wed Jun 12 2002 - 04:11:20 CDT

Hi again!

Today I have a new problem with the same protein-ligand-watercap system. I
have decided not to use any counter ions but to leave the protein charged
3+. I try to minimise the system, keeping everything fixed except water
molecules and hydrogens. I seem to have a problem with the keyword igb.
My input file looks like this:

  minimization
 &cntrl
    irest=0,ntx=1,
    imin=1,maxcyc=100,drms=0.001,
    temp0=300.0,ntt=1,tautp=0.2,
    ntc=1,ntf=1,
    nsnb=40,cut=15.0,
    dielc=1.0,
    ntpr=10,ntwx=0,ntwv=0,ntwe=0,
    ntb=0,ntp=0,taup=0.2,
    ipol=0,igb=0,
    scnb=2.0,scee=1.2,
    ncyc=10,ntmin=1,dx0=0.01,dxm=0.5,
    ibelly=1,
 &end

GRP1
ATOM 2 4
ATOM 6 6
ATOM 8 10
...
ATOM 4935 15269
END
END

If I use igb=0, which is default and should be ok since I have explicit
water molecules I get the following error message:

MPI: MPI_COMM_WORLD rank 0 has terminated without calling MPI_Finalize()
MPI: aborting job
 Program error

The output file just includes the input file, and not the last few rows of
it. I have tried to use both version 6 and 7 which gives the same error
message.
Do you have any clue what is wrong?
I tried to use other options for igb but nothing works:
1: not enough memory
2, 3: When igb>0 the moving part must be at the start of the molecule.

Is there anything wrong with my input? What can I do to make this
calculation run? Why cant I use igb=0?

Best regards,

        Emma Sigfridsson