Unable to use SA

Sergei V. Strelkov sergeis at DASH.CC.PURDUE.EDU
Sun Aug 3 18:03:33 EST 1997


Ravichandran N Venkataramani wrote:
> 
> Hi,
>   I am having a problem using simulated annealing... I am getting an error message
>   that read like this.... The r-factors start at (rf = 40, rw = 37) and go up to
>   (rf, rw = 55)... And the temperature and the energies in the verlet initations reach
>   exponential amounts (e+11)...
> 
>  %atoms "PRO2-340 -ALA -H   " and "PRO2-352 -VAL -CG2 "(XSYM#  4) only  0.76 A apart
>  %atoms "PRO2-340 -ALA -C   " and "PRO2-348 -GLY -N   "(XSYM#  4) only  1.48 A apart
>  NBONDS: found    59580 intra-atom interactions
>  NBONDS: found    32100 symmetry-atom interactions
>  XRTEST: number of selected reflections   11274
>  XRFILL: #scatt.= 2194 #anomalous=  0 #special pos.=  0 occupancies=1
>  XFFT: using grid [ 72,192, 48] and sublattice [ 72( 73),192(193), 48]
>  TRRESI: ->[TEST SET (TEST=1)] Fobs/Fcalc scale=   0.110 R=       0.542
>  TRRESI: ->[WORKING SET (TEST=0)] Fobs/Fcalc scale=   0.111 R=       0.534
>  %ATMCHK-ERR: unknown coordinates for atom "PRO1-59  -ARG -HH12"
>  %ATMCHK-ERR: unknown coordinates for atom "PRO2-226 -GLN -H   "
>  %ATMCHK-ERR: Unknown coordinates
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>  BOMLEV=    0 reached.  Program execution will be terminated.
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>  Subroutine DIE called . Terminating
> 
>    I have tried using the parameters nbond repel statement to minimize clash... But I
>  still get this error message....
> 
>  Moreover, when the model is viewed using O, I do not see any major clashes between
>  the atoms that Xplor has a problem with...
> 
>  Also, the model geometry is fine when analyzed using procheck...
> 
>  This problem has been baffling me for quite sometime... Any insights that you can
>  provide will be much appreciated
> 
>  ThankX-a-ton,
> 
>  Ravi.
>  venkatar at mail.med.upenn.edu

Ravi,

It's a typical blow-up of a SA run.

1. If some atoms are too close to each other 
(i.e. you are getting messages like
%atoms "PRO2-340 -ALA -C   " and "PRO2-348 -GLY -N   "(XSYM#  4) only 
1.48 A apart )
even before you start your SA run, you should probably manually correct
this.

2. If the clushing only appears in the middle of your SA run and just
before the blow-up,
_try decreasing your starting temperature_!
Note that, during the course of a SA run, R-factors normally do go up
initially 
-- because you run dynamics. The temperature will also go up a little
bit at the beginning,
and even _some_ clashing may be tolerated.

Good luck,
-- 
Sergei Strelkov                          Ph: (765) 494 4908 or 
Dept. of Bio. Sciences, 1392 Lilly Hall,           494 6766 (lab) 
Purdue University, West Lafayette,           (765) 743 8405 (h)    
IN 47907-1392, USA.                     Fax: (765) 496 1189



More information about the X-plor mailing list