AMBER: bad nonbon in nmode (amber9)

From: Enghui Yap <enghui.berkeley.edu>
Date: Mon, 14 Jul 2008 09:25:29 -0700

Hi

I am trying to do normal mode analysis on a protein with 3906 atoms, in
vacuo. I have minimized the x-ray structure using sanders to drms<1e-5. To
 use nmode, I change the maxpr in nonbon1 from 2800 to 3910.

When I tried do further minimization in nmode using newton-raphson (ntrun=4) I get the
following error:

          rms of step length = 0.0000000629
Bad nonbon: 1 18 0.76921E+00 0.84351E+03
Bad nonbon: 1 21 0.30746E+01 0.50627E+03
...
Bad nonbon: 1899 1914 0.29739E+01 0.54335E+02
Bad nonbon: 1899 1915 0.69180E+00 0.66862E+02
          ene .step k-1 = -7952.339035 ene @k = ***************
          ... scaling was unsuccessful ... stop

Here's my nmode input file:
&data
   ntrun = 4,
   maxcyc = 200,
   drms = 0.000001,
   cut = 999,
   idiel = 1,
   scee = 1.2,
   nprint = 5,
   nvect = 0,
   ntx = 0,
   ntxo = 0,

   smx = 0.01,
   alpha = 0.5,

/

I've checked that the energies from the final step of sander's
minimization is the same as the first step of nmode's minimization:
-----------------------------------------------------------------
(Sanders)
                     FINAL RESULTS
   NSTEP ENERGY RMS GMAX NAME NUMBER
  42217 -7.9523E+03 7.8954E-06 1.2393E-04 HA 2529
 BOND = 149.7237 ANGLE = 466.9506 DIHED = 2412.6363
 VDWAALS = -2102.6763 EEL = -19186.0876 HBOND = 0.0000
 1-4 VDW = 828.1526 1-4 EEL = 9478.9616 RESTRAINT = 0.0000
-----------------------------------------------------------------
(nmode)
   step = 0
     F = -0.795234E+04 GRDMAX = 0.123933E-03 GNORM = 0.789669E-05
       E-NONB E-ELE E-HBOND E-BOND
    -0.21027E+04 -0.19186E+05 0.00000E+00 0.14972E+03
       E-ANGLE E-DIHED E-NB14 E-EEL14
     0.46695E+03 0.24126E+04 0.82815E+03 0.94790E+04
       E-POL E-3BOD
     0.00000E+00 0.00000E+00
-----------------------------------------------------------------

I tried printing out the energies in forces.f and enb and eel are
undefined (*****)

I think that the error is due to some atom pairs being too close in
nonbon1, but I am not sure how that happen, assuming the xdir is
calculated correctly? I've stared at the code and am still baffled. any
help is appreciated!

Thanks,
Enghui

p/s: I managed to use the restart file from Sander's minimization directly for normal mode analysis (ntrun=1) and it ran ok. But I'd still like to understand why ntrun=4 did not work.

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo.scripps.edu
Received on Wed Jul 16 2008 - 06:07:21 PDT
Custom Search