Previous Thread
Next Thread
Print Thread
Randomly crashing CPT, constrained simulation.
#25219 08/28/10 08:13 PM
Joined: Jan 2004
Posts: 4
P
Forum Member
OP Offline
Forum Member
P
Joined: Jan 2004
Posts: 4
I have a problem with a 5nm3 simulation crashing almost randomly after 1-3ns.

The simulation contains one 26 amino acid peptide, with its backbone constrained.
Some 100 sugars, and a few thousand waters (TIP3P)
Timestep is 2fs
The job is run with charmm 34b1 using MPIrun on 8 processors.

The simulation runs fine, but then randomly crashes after 1-3 ns.
Crash line is 'Energy Tol. Exceeded'

The bizarre thing is the simulation is apparently running fine.
Everything is moving as it should. There is no apparent significant change in the energy or temperature of the system.
Boundary conditions look fine. Box size is right on the money.

The input file is below:
!--- topology and parameter files
open read card unit 10 name "top_all22_prot.inp"
READ RTF CARD UNIT 10
CLOSE UNIT 10





OPEN READ CARD UNIT 10 NAME "par_all22_prot.inp"
READ PARAMETERS CARD UNIT 10
CLOSE UNIT 10


!--- Protein part ----
READ SEQUENCE CARD
*sequence for 2MLT
*
26
GLY ILE GLY ALA VAL LEU LYS VAL LEU THR THR GLY LEU
PRO ALA LEU ILE SER TRP ILE LYS ARG LYS ARG GLN GLN

GENERATE PROT LAST CT2 angles dihedrals setup

READ SEQUENCE CARD
*sequence for IONS
*
6
CHLR CHLR CHLR CHLR CHLR CHLR

GENERATE IONS angles dihedrals setup

READ SEQUENCE CARD
*sequence for IONS
*
120
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC
AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC AGLC


GENERATE GLUC angles dihedrals setup

!patch beta GLUC 21 setup
set a = 1

label loop

patch beta GLUC @a setup
incr a by 1
if a NE 121 goto loop


READ SEQUENCE TIP3 2406
GENERATE WAT NOANGLE NODIHE

define water select segid wat end



!therm was 10ps @ eq every 60
OPEN UNIT 4 READ CARD NAME "start_coord_beta.crd"
READ COOR CARD UNIT 4


SET 9 46.46538

CRYSTAL DEFINE CUBIC @9 @9 @9 90 90 90
CRYSTAL BUILD CUTOFF 14 NOPERATIONS 0


update CUTNB 13.5 CTONNB 10.5 CTOFNB 11.5 cutim 13.5 VSWITCH SHIFT


IMAGE BYRES XCEN 0.0 YCEN 0.0 ZCEN 0.0 SELECT segid wat END
IMAGE BYRES XCEN 0.0 YCEN 0.0 ZCEN 0.0 SELECT segid GLUC END
IMAGE BYSEG XCEN 0.0 YCEN 0.0 ZCEN 0.0 SELECT segid PROT END

define backbone sele type N .or. type CA .or. type C end
!cons harm absolute sele backbone end

cons harm main forc .500 sele backbone end


shake bonh param

UPDATE CUTIM 13.5
!Mini SD NSTEP 200

OPEN WRITE CARD UNIT 20 NAME "betagluc_melittin_cpt.rst"
OPEN WRITE FILE UNIT 21 NAME "betagluc_melittin_cpt.dcd"
OPEN WRITE CARD UNIT 22 NAME "betagluc_melittin_cpt.vel"

Dyna cpt -
Strt NSTEP 25000000 TIMESTEP 0.002 -
FIRSTT 300.0 FINALT 300.0 TSTRUCT 300 -
TWINDL -10.0 TWINDH 10.0 IEQFRQ 0 IHTFRQ 0 -
IASORS 0 IASVEL 1 -
NTRFRQ 50000 echeck 9999 -
EWALD PMEWALD KAPPA 0.3 ORDEr 6 FFTX 48 FFTY 48 FFTZ 48 -
NSAVV 2000 NSAVC 2000 IUNCRD 21 IUNWRI 20 IUNREAD 23 -
KUNIT 22 NPRINT 1000 -
ISVFRQ 1000 -
PCONST PINT PREF 1.0 PMASS 2000.0 pgamma 20 -
HOOVER REFT 300.0 TMASS 10000.0 -
CDIE EPS 1.0 -
imgfrq -1 inbfrq -1 nbxm 5 e14fac 1.0

OPEN UNIT 8 WRITE FORM NAME "beta_end.crd"
WRITE COOR CARD UNIT 8

OPEN UNIT 8 WRITE FORM NAME "beta_end.pdb"
WRITE COOR pdb UNIT 8

STOP
----------------------------

Im kinda pulling my hair out on this. The simulation looks like it runs fine for about 1-2 days, then just dies with this energy tolerence exceeded thing.

Ive tried lots of tinkering with the CPT stuff, but nothing seems to solve it (although it does change the step on which the simulation crashes.

While I suspect its some funny interplay between the constraint and the CPT, I would appreciate anyone who can help me getting this thing stable,
Best wishes,
Phil Mason - Brady lab.

Re: Randomly crashing CPT, constrained simulation.
Proteios #25220 08/28/10 08:47 PM
Joined: Sep 2003
Posts: 8,501
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,501
You should not use either CONS FIX or CONS HARM ABSOLUTE with constant P; the coordinates are scaled along with the box edges.

To preserve the backbone shape, a number of other restraint methods can be used with CPT, including CONS DIHE, CONS IC, CONS RMSD, or maybe CONS HARM BESTFIT applied to residue ranges. (Note that, for backwards script compatibility, CONS HARM without either BESTFIT or RELATIVE keywords implies ABSOLUTE by default.)


Rick Venable
computational chemist


Moderated by  lennart, rmv 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.4
(Release build 20200307)
Responsive Width:

PHP: 5.6.33-0+deb8u1 Page Time: 0.006s Queries: 18 (0.003s) Memory: 0.8974 MB (Peak: 0.9622 MB) Data Comp: Off Server Time: 2020-10-21 08:02:13 UTC
Valid HTML 5 and Valid CSS