Problem for allocating space

Ezra Peisach epeisach at AURIGA.ROSE.BRANDEIS.EDU
Thu Nov 13 10:34:33 EST 1997


What this means is that the machine you are on did not have the memory
available to give you another 3MB of memory. 

Depending on the machine there are several options:

a) Check to make sure that you have not gone over the per process
memory allocation. 

If you type "limit" you will see some per process memory limits.
Type "unlimit" and the allowed memory use will be increased, as well as
your allowed virtual memory use.

Try again....

b) If that is not the problem: See if the machine you are on has run out
of swap space - your system administrator can probably help. If so,
the admin can increase the swap space or maybe free up some memory by
getting rid of extraneous processes. - i.e. exit all xterms, netscape, etc
that are not needed.

c) In the fft stanza in xplor, change the memory statement.... This may
have adverse affects on your computation - i.e. they may take alot
longer...

Good luck.

	Ezra

 On 13 Nov 1997 guoguang at alfa.mbb.ki.se wrote:

> Hi, X-PLOR users
> 
> I am trying to do a bulk solvent correction to a fairly big protein (total 
> 2200 residues in an assymetric unit) and with 160,000 reflections. However
> the program (xplor 3.851) crashed with following message
> 
>  XREFINE>   do (fpart=ft(mask)) ( all ) 
>  XMAPAL: allocating space for real space object.
>  Total of   2880000 map elements were selected.
>  %VEHEAP-ERR: could not satisfy storage request for     2916002 bytes
>  Subroutine DIE called . Terminating
> 
> How should I cope this? 
> 
> thanks in advance for any hints
> 
> Guoguang
> 
> p.s: my input file
>  remarks  file xtalrefine/setup_bulksol.inp 
>  remarks
>  remarks  Generates a "flat" bulk solvent model
>  remarks 
>  remarks  Computes solvent mask, do an FFT, optimize k and B, stores in FPART
>  remarks  and writes new reflection file. 
>  remarks
>  remarks 
>  remarks   - run this job
>  remarks   - check that the solvent parameters are reasonable (solvent density 
>  remarks     and B-value) and that there is a significant drop in R value 
>  remarks     especially at low resolution.   If unsatisfactory, try
>  remarks     fixing either the solvent B-factor or solvent density level. 
>  remarks   - use setup_bulksol.cv reflection file in all subsequent jobs
>  remarks   - use lower resolution bound in all subsequent jobs
>  remarks  Authors: Jian-Sheng Jiang, William I. Weis and Axel T. Brunger
>  
> 
>  ! Cite the following references when using this protocol:
>  !
>  ! J.-S. Jiang and A.T. Brunger, Protein hydration
>  ! observed by x-ray diffraction: solvation properties of
>  ! penicillopepsion and neuraminidase crystal structures,
>  ! J. Mol. Biol. 243, 100-115 (1994).
>  !
> 
> 
> !----------------------------------------------------------------------------
> 
>  
> {===>} parameter 
>  @/nfs/disk1/guoguang/xplor/parhcsdx.pro
> @/nfs/disk1/guoguang/pdc/par.tdp
> @/nfs/disk1/guoguang/pdc/par.mg
> @par.cit
> @/nfs/disk1/guoguang/xplor/param19.sol   { Read empirical potential       }
> end           {*Read parameters.*}
> 
> {===>} structure @gen.psf end       {*Read structure file.*}
> 
> {===>} coor @gen.pdb                              {*Read coordinates.*}
> 
> 
> {===>} evaluate ($scatter_library="/usr/programs/xplor3.84/xtallib/scatter.lib")  
>                                                      { Form factor library. }
>                                              
> {===>}                { Space group. Uses International Table conventions.  } 
>                       { with supercripts substituted by parenthesis.        }                                            
>  evaluate ($SG="P1" ) 
> 
> {===>}                                                         { Unit cell. }
>  evaluate ($a=69.87)
>  evaluate ($b=92.01)
>  evaluate ($c=97.99)
>  evaluate ($alpha=103.68)
>  evaluate ($beta=94.54)
>  evaluate ($gamma=112.32)
> 
> 
> {===>}
>  evaluate ($ref="/nfs/disk2/doreen/cof/p1cof2c.cv")   { Input reflection file. 
> Includes}
>                                            { test set for cross-validation. }
>  
> {===>}
>  evaluate ($out_ref="/nfs/scratch/guoguang/bulksol.cv")    { Output reflection 
> file.        }
> 
> {===>}
>  evaluate ($low_res=15.0)                       {* low resolution limit.  *}
>  evaluate ($high_res=1.86)                        {* high resolution limit. *}
>  
> {===>} 
>  evaluate ($f_cut=1.0)                        {* F/sigma amplitude cutoff. *}
> 
> {===>}    
>  evaluate ($f_low=0.00)                    {* Absolute amplitude cutoffs. *}
>  evaluate ($f_high=1000000) 
> 
> 
> 
> {===>}                       {* Select atoms to be included in refinement. *}
>  vector ident (store1) (known) 
> 
> {===>}
>  evaluate ($ncs_flag=NONE) { RESTRAIN or STRICT or NONE ! ncs information.  }
>  
> {===>}
>  evaluate ($ncs_file="ncs.inp" )  { NCS-restraints/constraints file name.        
>   }
>                            { Examples are in xtalrefine/ncs_strict.inp      }
>                            { and xtalrefine/ncs_restrain.inp.               }
>  
> {===>}
>  evaluate ($k3=0.33) {* Optional: fix set solvent density level by setting}
>                        {* this parameter to a positive value.               }
>                        
> {===>}
>  evaluate ($b3=50.) {* Optional: fix set solvent b-factor by setting     }
>                        {* this parameter to a positive value.               }
>            
> !----------------------------------------------------------------------------
> 
>  xrefine
> 
>    @/usr/programs/xplor3.84/xtallib/spacegroup.lib       { Read symmetry 
> library. }
> 
>    a=$a b=$b c=$c  alpha=$alpha beta=$beta gamma=$gamma { Define unit cell. }
>   
>    @$scatter_library                             { Read form factor library }
> 
>    nreflections= 160000
>    reflection @$ref end
> 
>    resolution_limits= $low_res $high_res
> 
>    do (fobs=0) (amplitude(fobs) <= $f_cut * sigma) 
> 
>    fwindow $f_low  $f_high    
> 
>    method=FFT          
>    
>    fft 
>       memory=3000000   
>    end                 
>    
>  
>    selection=( store1 ) 
> 
> end
> 
>  if ($ncs_flag=STRICT) then
>     @@$ncs_file
>  elseif ($ncs_flag=RESTRAIN) then
>     @@$ncs_file
>     flags include ncs end
>  end if
>               
> xrefin
>    
>    update                        {*Update Fcalcs and print current rfactor.*}
>    
>    print rfactor
>    
> end                  
> 
> 
>  xrefine
>   
>    {* compute a solvent mask and store it in mask *}
>    declare name=mask domain=real end
>    mask               
>      mode=vdw  		{ use vdw radii                     }
>      solrad=0.25        { a probe radius adds to vdw radii  }
>      shrink=0.25        { the shrink radius                 }
>      nshell=1           { the number of shells              }
>      to=mask            { output to the mask map            }
>    end
>    
>    {* Fourier transformation of the solvent mask and store it in fpart*}
>    do (fpart=ft(mask)) ( all )
> 
>    {* solvent parameters refinement via by the multiscale routine *}
>    multiscale
>       bfmin=0. bfmax=200.
>       set1=fobs   k1=-1       b1=0
>       set2=fcalc  k2=1.0      b2=0 
>       set3=fpart  k3=$k3      b3=$b3        
>       selection=( $high_res <= d <= $low_res 
>              and  $f_low <= amplitude(fobs) <= $f_high
>              and test=0 )
>       resk=5.0             { resolution boundary for two FFKs }
>       update=false
>       bmin=10.
>       bmax=300.
>       
>       ?
>    end
> 
>    {* compute refined solvent structure factors in fbulk *}
>    do (fpart=$k3*exp(-$b3*s()*s()/4.0)*fpart) ( all )
>    
>    display  bulk solvent model parameters: density level = $k3 e/A^3,  B-factor= 
> $b3 A^2
>    remarks  bulk solvent model parameters: density level = $k3 e/A^3,  B-factor= 
> $b3 A^2
>    
>    print rfactor
>    
>    do (fcalc=complex(0,0)) ( all )
> 
>                   {*Write a new reflection file including the solvent FPART.*}
>    write reflection 
>       output=$out_ref sele=( all ) 
>    end   
> 
>  end
> 
>  stop
> 
> 




More information about the X-plor mailing list