AMBER Archive (2004)

Subject: RE: AMBER: Sander : Vrand flag

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Mon Aug 23 2004 - 15:58:15 CDT

Dear Sachin,
 
For a non-periodic system there is no easy way to control the density of the
system since there is no real concept of pressure in this situation.
However, your solvent cap artifically creates a force that holds the solute
in place and so could be thought of as applying pressure ot the system. In
order to get a density of 1g cm-1 you will need to vary the size of your
solvent cap. Note, however, that the density of bulk water at standard
temperature and pressure is 1g cm-1. What you have is a protein in a very
small droplet of water. In this situation the density is definately not 1g
cm-1. Your best option is to add your own water molecules and then use the
TIP3PBOX (WATBOX216) water system with the solvate cap command. This will
add an equilibrated set of waters around your system that have a reasonable
density. Then you can equilibrate your system. If you find that the density
seems too small you can then vary the cap dimensions and equilibrate again.
Ideally you should let the system equilibrate to a density it is stable in,
this is only possible with constant pressure periodic boundaries, however.
In you situation the best you can do is ensure you have a reasonable number
of waters within the volume defined by your cap.
 
Note, if the initial density of your system is not uniform it should become
uniform after sufficient equilibration.
 
I hope this helps.
All the best
Ross
 
/\
\/
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk/ | PGP Key available on request |

 
 

  _____

From: owner-amber_at_scripps.edu [mailto:owner-amber_at_scripps.edu] On Behalf Of
sachin patil
Sent: 23 August 2004 12:55
To: amber_at_scripps.edu
Subject: RE: AMBER: Sander : Vrand flag

Dear Dr. Ross,
 Thank you for all the informatiom you are providing, that too in real quick
time.
We really appreciate that.
 One last question is, is it possible to control the density of the system,
so that after adding water molecules, I can make the density unifom (
density of water is 1)
 I am using Amber7.
 
 
Thank you again,
 
Regards,
Sachin Patil,
============
Graduate student,
Dept. of Medicinal & Biological Chemistry
College of Pharmacy
University of Toledo,
Toledo, Ohio 43606
 

Ross Walker <ross_at_rosswalker.co.uk> wrote:

Dear Sachin,
 
A couple of things to try. It is possible, although I am not sure, that the
extra water molecules you are adding are not subject to the constraint force
of the solvent cap and so can boil off into space. Adding these extra waters
is also likely to greatly increase the density at the regions in which they
are being added leading to large initial forces. If you think about it you
have a droplet of water at a reasonable density to which you are adding
extra waters. Hence the effective "pressure" inside the droplet is
increasing significantly. I would try adding your extra waters BEFORE you
add the solvent cap. In this way you won't increase the density since the
solvatecap command will add less waters to compensate.
 
The reason the temperature is going up so high is that your system is
exploding. The first sign of an "exploding" system is when the temperature
rapidly shoots up. The temperature coupling is only weak so doesn't have
time to 'correct' for this sudden increase in kinetic energy.
 
As for the vrand setting I would leave this at zero since it will greatly
effect the dynamics of your system and mask what you are trying to measure.
I would avoid ntt=2 for the moment and try and see if you get stable
dynamics using a simple Berendsen thermostat. Set ntt=1 and tautp to
something reasonably small like 1.0ps. Then I would take you solvent cap
system and minimise if for quite a bit (Ca 5000 steps) before you move to
any MD. Then try setting tempi=0 and temp0=300 and let your system slowly
heat up. You could run say 20ps of MD here and dump the coordinates every
step. In this way you will be able to visualise the fine dynamics of your
system and if it blows up you will be able to see from where this happens
and hopefully have a better chance of fixing it.
 
I would also try increasing the force constant of the water cap (FCAP) above
the default (which I think is 1). Note you may also need to manually turn on
the water cap in Amber 7. In amber 8 it is turned on by default if cap
information is detected in the prmtop file. This may not be true for Amber 7
- I would check your output carefully.
 
All the best
Ross
 
/\
\/
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk/ | PGP Key available on request |

 
 

  _____

From: owner-amber_at_scripps.edu [mailto:owner-amber_at_scripps.edu] On Behalf Of
sachin patil
Sent: 23 August 2004 12:08
To: amber_at_scripps.edu
Subject: RE: AMBER: Sander : Vrand flag

Dear Dr. Ross Walker,
 I am trying to visualize movement of water molecules and their involment in
the ligand binding. I have a receptor and ligand system which is around 6500
atoms big. I solvated it with a watercap and then manually added few water
molecules in the system.
 When I load this modified pdb file in Xleap and make a "check" it gives me
couple of error messages one of which is for water molecules.
I have restrained a few hydrogen bonds and the receptor to it position (
force constant 2). I am running nonperiodic conditions.I had set the
temperature to be 298 K, but during the course of simulation the
temperatures go above 350 K and the water molecules boil.
 
 (Interestingly, with another system which has no manually added water
molecules the simulation goes fine. Eventhough the temperatures and high,
similar to above.)
 
Since I just wanted the surface of interst, i.e. ligand binding site to be
solvated so I added a water cap. If I solvate the entire system (so as to
run periodic bounadry condition the system will be huge.)
 And since I wanted to control the temperatures, I set vrand=1 ( since i
couldnot use Longevian alorithm ntt=3, with amber7.) and it took the
simulation through.
 Do you still suggest me to solvate the entire system and then run periodic
boundary
coinditions after manually adding water molecules?
 
 Or shall I check the system for any close contacts.
 In that case can you please brief me how the detalied initial minimization
should be.As if now, I first minimize the water molecules and then the
entire system.
 I have attached the input and sample output files below.
 
Thank You
 
Regards,
Sachin Patil,
============
Graduate student,
Dept. of Medicinal & Biological Chemistry
College of Pharmacy
University of Toledo,
Toledo, Ohio 43606
 
Follwing is the sample out put-
 
 =======================================================
 
Setting new random velocities at step 8500
 NSTEP = 8500 TIME(PS) = 218.500 TEMP(K) = 149.65 PRESS = 0.0 Etot =
-17275.3978 EKtot = 3503.8761 EPtot = -20779.2738 BOND = 1126.2674 ANGLE =
3085.7870 DIHED = 3889.4866 1-4 NB = 1416.2147 1-4 EEL = 13519.4021 VDWAALS=
-1183.8538 EELEC = -43371.0430 EHBOND = 0.0000 RESTRAINT = 738.4651 EAMBER
(non-restraint) = -21517.7389
---------------------------------------------------------------------------- 

--
----------------------------------------------------------------------------
--
 NSTEP = 8550 TIME(PS) = 218.550 TEMP(K) = 300.98 PRESS = 0.0 Etot =
-11965.3497 EKtot = 7046.9539 EPtot = -19012.3036 BOND = 1099.3445 ANGLE =
3024.5770 DIHED = 3922.5966 1-4 NB = 1448.0821 1-4 EEL = 13583.3557 VDWAALS=
-1153.0540 EELEC = -41662.9215 EHBOND = 0.0000 RESTRAINT = 725.7160 EAMBER
(non-restraint) = -19738.0196 
--------------------------------------------------------------------------
-------------------------------------------------------------------------
 
 Here is the input file:
 
 &cntrl nstlim=190000,dt=0.001, 
cut=12,ntb=0,nscm=1000, temp0=298,
tempi=298, vrand=100,ntt=1, 
irest=1,ntx=5,nmropt=1, ntr=1,tautp=2.0,
fcap=.5, ntf=2, ntc=2,
ivcap=1,ntwe=500, ntwx=500, ntpr=50, 
ntwr=5000, nrespa=1,matcap=6178, &end 
&wt type='END' 
&end 
DISANG=./distfinal.rst 
LISTOUT=POUT 
group input for restrained atoms 
2.0 
RES 1 375 
END 
END 

Ross Walker <ross_at_rosswalker.co.uk> wrote:

Dear Sachin
 
By setting vrand=100 you are randomising the velocities every 100 steps.
This is probably not a good idea for the problem you are trying to solve.
The fact that your system is exploding suggests that something is wrong with
it. How did you create it? You probably have some very close contacts
somewhere.
 
You also appear to be running a vacuum simulation but you talk about waters?
How did you solvate your system? Did you add a solvent cap and a restraint
force on the cap? If not then your system is likely to just boil off into
space. Ideally if you want to run a solvated system you should use periodic
boundaries.
 
I would try running extensive minimisation of your system first. Then try
heating it slowly (possibly with a time step as short as 0.5fs if you have
to) from 0 to 100K or so over 20ps at constant volume, assuming you can run
periodic boundaries. If it is stable here then try slowly increasing the
temperature over another 20ps or so to 300K. Then equilibrate for a while at
300K before switching to constant pressure.
 
If you still have problems please post details about your system as well as
the input and output files.
 
All the best
Ross
 
/\
\/
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk/ | PGP Key available on request |

  _____  

From: owner-amber_at_scripps.edu [mailto:owner-amber_at_scripps.edu] On Behalf Of
sachin patil
Sent: 23 August 2004 11:00
To: amber_at_scripps.edu
Subject: AMBER: Sander : Vrand flag 

Dear Amber Users, 
In the MD simulations I am curently running, the system was constanly
exploding and I could see few water molecules out of the systems. And was
getting an error in the MDOUT file as-
The system has extended beyond the extent of the virtual
box..................
.....restarting shold resolve the error.
And the tempratures were constantly getiing higher and higher.
 
But when I set vrand=100 the simulation went throught just fine. But now my
problem is, the temperature of the system is too cold after every 100 steps.

 
I dont know the implications of this on my system.And I am not sure what
effect this has on the final results. 
So any experiences on similar problems, answers, suggestions or comments are
wel-comed.
 
Thank You,
 
Sachin Patil,
============
Graduate student,
Dept. of Medicinal & Biological Chemistry
College of Pharmacy
University of Toledo,
Toledo, Ohio 43606
 
Follwing is the sample out put-
 
 ======================================================= 
 
Setting new random velocities at step 8500 
 NSTEP = 8500 TIME(PS) = 218.500 TEMP(K) = 149.65 PRESS = 0.0 Etot =
-17275.3978 EKtot = 3503.8761 EPtot = -20779.2738 BOND = 1126.2674 ANGLE =
3085.7870 DIHED = 3889.4866 1-4 NB = 1416.2147 1-4 EEL = 13519.4021 VDWAALS=
-1183.8538 EELEC = -43371.0430 EHBOND = 0.0000 RESTRAINT = 738.4651 EAMBER
(non-restraint) = -21517.7389 
----------------------------------------------------------------------------
--
----------------------------------------------------------------------------
--
 NSTEP = 8550 TIME(PS) = 218.550 TEMP(K) = 300.98 PRESS = 0.0 Etot =
-11965.3497 EKtot = 7046.9539 EPtot = -19012.3036 BOND = 1099.3445 ANGLE =
3024.5770 DIHED = 3922.5966 1-4 NB = 1448.0821 1-4 EEL = 13583.3557 VDWAALS=
-1153.0540 EELEC = -41662.9215 EHBOND = 0.0000 RESTRAINT = 725.7160 EAMBER
(non-restraint) = -19738.0196 
--------------------------------------------------------------------------
-------------------------------------------------------------------------
 
 Here is the input file:
 
 &cntrl nstlim=190000,dt=0.001, 
cut=12,ntb=0,nscm=1000, temp0=298,
tempi=298, vrand=100,ntt=1, 
irest=1,ntx=5,nmropt=1, ntr=1,tautp=2.0,
fcap=.5, ntf=2, ntc=2,
ivcap=1,ntwe=500, ntwx=500, ntpr=50, 
ntwr=5000, nrespa=1,matcap=6178, &end 
&wt type='END' 
&end 
DISANG=./distfinal.rst 
LISTOUT=POUT 
group input for restrained atoms 
2.0 
RES 1 375 
END 
END 

  _____  

Do you Yahoo!?
Yahoo!
<http://us.rd.yahoo.com/mail_us/taglines/50x/*http://promotions.yahoo.com/ne
w_mail/static/efficiency.html> Mail - 50x more storage than other providers!

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

  _____  

Do you Yahoo!?
Read only the mail you want - Yahoo!
<http://us.rd.yahoo.com/mail_us/taglines/spamguard/*http://promotions.yahoo.
com/new_mail/static/protection.html> Mail SpamGuard.

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu