RE: AMBER: problem in implcit simulation

From: Ross Walker <ross.rosswalker.co.uk>
Date: Wed, 28 Mar 2007 22:02:28 -0700

Dear Gurpreet,
 
Your assumption that implicit solvent simulations should be faster than
explicit solvent simulations is not strictly true. For small numbers of
atoms it is certainly true but as you move to larger and larger systems so
you reach a cross over point where implicit solvent becomes faster. Where
this crossover occurs depends on a number of factors including the size of
the cutoffs you use, how much water you have in the explicit solvent case,
which GB options you are using and even the hardware you are running on
etc...
 
Typically I find for a GB cutoff of around 18 angstroms vs explicit solvent
with 8 angstrom cutoff on a P4 the crossover is around 2,000 to 3,000 atoms
although you may get different numbers. If you are running say a 20,000 atom
system then the GB simulation will be much much slower.
 
The reason for this is that the GB equation is much more complicated than
the direct space + PME equation of explicit solvent. Similarly it scales
worse, N^2 if you don't use a cutoff hence as you go to larger systems it
becomes progressively more expensive.
 
 
All the best
Ross
 
/\
\/
|\oss Walker

| HPC Consultant and Staff Scientist |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
| http://www.rosswalker.co.uk <http://www.rosswalker.co.uk/> | PGP Key
available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.
 



  _____

From: owner-amber.scripps.edu [mailto:owner-amber.scripps.edu] On Behalf Of
gurpreet singh
Sent: Wednesday, March 28, 2007 21:26
To: amber.scripps.edu
Subject: AMBER: problem in implcit simulation


hello Amber users

i am using sander module of Amber9 to carry out the implcit MD simulation
for my protein , since i m new to implicit simulations so i dont know
exactly the time scales of these simulations but one thing which i am very
sure about is that implicit shud take less time as compared to the explicit
simulation but in my case opposite is happening,implicit mini is taking lot
more time as comapred to explicit, i started my simulation with the
minimization by putting restraint on all the heavy atoms and allowed
hydrogen atoms to move freely, same rpoblem i observed in equilibration run
which i carried after this mini

i m giving my mini file flags please tell me that this problem of more time
is comin because of some particular flag i mnetioned in the input file or
because of some other problem

mini file : imin=1,igb=2,cut=20.0,maxcyc=500,ncyc=250,ntr=1,intdiel=1.0
extdiel=78.5,ntb=0


equilibration file: imin=0,igb=2,cut=20.0,ntr=1,intdiel=1.0
extdiel=78.5,ntb=0,nstlim=20000,
dt=0.002,ntc=2,ntf=2,tempi=10.0,temp0=300.0,ntt=1,tautp=2.0,ntp=0,ntpr=100,
ntwx=0

please reply soon
 thanks



Gurpreet Singh
IIT MADRAS
INDIA




-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Apr 01 2007 - 06:07:20 PDT
Custom Search