towhee-users Mailing List for MCCCS Towhee

On Sep 29, 2005, at 5:52 AM, Huiqun Zhou wrote:
> I have succeeded in compiling mpitowhee and towhee on a P4 cluster and=20=

> an EM64T machine,
> but failed in creating the two programs aforementioned, respectively.=20=

> I tried with and without
>  optimization,=A0got same error when linking the programs:
> =A0=A0 ....../crt1.o (.text+0x18): In function _start
> =A0 : undefined reference to 'main'
> =A0
> Could anyone give me a pointer how to compile the programs correctly?

Make sure you have cleaned out all of the old object files when you go=20=

to compile these versions.

Starting in the base directory
./configure FFLAGS=3D-fbounds-check
cd Source
make clean
make forcefield
make fitcoex

should do the trick.

Marcus
-------------------------------------------------------------------
Marcus G. Martin
Senior Member of Technical Staff
Multiscale Computational Materials Methods
Sandia National Laboratories
PO Box 5800 MS 1110
Albuquerque NM 87185-1110
v. 505 284-6355
f. 505 845-7442
http://www.cs.sandia.gov/marcusmartin/
http://towhee.sourceforge.net

dear list users,

I have succeeded in compiling mpitowhee and towhee on a P4 cluster and =
an EM64T machine,
but failed in creating the two programs aforementioned, respectively. I =
tried with and without=20
optimization, got same error when linking the programs:
   ....../crt1.o (.text+0x18): In function _start
  : undefined reference to 'main'
=20
Could anyone give me a pointer how to compile the programs correctly?

Thanks in advance.


Huiqun Zhou

Hello towhee-users,

There was a file missing from the recent Towhee version 4.11.6 release. 
  This is the new readutils.h file.  It is now available for download on 
the SourceForge website and will be included in future releases.  Place 
this file in the Source directory in order to compile the code 
properly.

Marcus

-------------------------------------------------------------------
Marcus G. Martin
Senior Member of Technical Staff
Multiscale Computational Materials Methods
Sandia National Laboratories
PO Box 5800 MS 1110
Albuquerque NM 87185-1110
v. 505 284-6355
f. 505 845-7442
http://www.cs.sandia.gov/marcusmartin/
http://towhee.sourceforge.net

On Sep 24, 2005, at 8:40 PM, li jianguo wrote:
> I want to define a new atom type to construct my molecule=A0using UFF.=20=

> I=A0defined the atom type and its corresponding parameters=A0in=A0the=20=

> forcefiled file: towhee_ff_UFF, but when runing the job, I got error=20=

> message: parameters not found for atom
>  Is there any other file should I modify=A0or shoud I recompile it=20
> before running the job?

If you modify the files properly it should work without having to=20
recompile.  The only time you would have to recompile is if you were=20
building your own ffwhatever.F file and using the forcefield utility to=20=

generate the towhee_ff_whatever file.

I would suggest creating a new forcefield file when you wish to add=20
your own parameters as then it will be much easier to upgrade to new=20
versions without worrying about the new changes having overwritten your=20=

own modifications.  Especially for UFF as I seem to be constantly=20
finding more bugs in that implementation as the paper describing that=20
force field leaves quite a bit as an exercise for the reader.

On Sep 25, 2005, at 9:34 PM, li jianguo wrote:
> =A0INVERTHMATRIX: zero boxvolume=A0=A0 0.
> =A0VOLNVT: move rejected, trial volume too low
> =A0VOLNVT: a trial volume went negative
> =A0should not happen in the NVT version of volume
> Could anybody tell me what is the reason and how to fix it?

Not from that limited information.  When you get messages such as these=20=

you should check to make sure you didn't set up something obviously=20
wrong.  If you can't find anything then please submit a bug report=20
(including input files) to towhee-bugs

On Sep 26, 2005, at 6:39 AM, li jianguo wrote:

> I want to=A0fix the bond and angle of a molecule by using UFF,=A0can=20=

> anybody kindly tell me=A0how to=A0achieve these in Towhee?

There is not currently any functionality in Towhee to fix bonds and=20
angles when the force field says they are flexible.  If you want a=20
rigid version of UFF you are going to have to create your own force=20
field file.  This will likely be very difficult for UFF given its=20
rather unique description of the intramolecular terms.

Marcus

-------------------------------------------------------------------
Marcus G. Martin
Senior Member of Technical Staff
Multiscale Computational Materials Methods
Sandia National Laboratories
PO Box 5800 MS 1110
Albuquerque NM 87185-1110
v. 505 284-6355
f. 505 845-7442
http://www.cs.sandia.gov/marcusmartin/
http://towhee.sourceforge.net

Hello towhee-users,

A couple of significant Towhee updates.

First, the project has now moved completely to the SourceForge website. 
  A notable change for the users is the release distributions are now 
available for download from the SourceForge website instead of through 
a Sandia website.

Second, the recently released version (hopefully) finally has a fully 
debugged version of the new configurational-bias Monte Carlo options 
that utilize arbitrary trial distributions.  If you have been using 
these features you will want to download the new version as there were 
several mistakes in the previous versions.

Marcus

-------------------------------------------------------------------
Marcus G. Martin
Senior Member of Technical Staff
Multiscale Computational Materials Methods
Sandia National Laboratories
PO Box 5800 MS 1110
Albuquerque NM 87185-1110
v. 505 284-6355
f. 505 845-7442
http://www.cs.sandia.gov/marcusmartin/
http://towhee.sourceforge.net

Dear all,
I want to fix the bond and angle of a molecule by using UFF, can anybody kindly tell me how to achieve these in Towhee?
Thanks in advance,
Li JG
Department of Chemical engineering
National University of Singapore 



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

Dear all,
I am running NVT Gibbs ensemble MC by Towhee, but often got error message:
...
INVERTHMATRIX: zero boxvolume   0.
 VOLNVT: move rejected, trial volume too low
 INVERTHMATRIX: zero boxvolume   0.
 VOLNVT: move rejected, trial volume too low
 INVERTHMATRIX: zero boxvolume   0.
 VOLNVT: move rejected, trial volume too low
 INVERTHMATRIX: zero boxvolume   0.
 VOLNVT: move rejected, trial volume too low
 INVERTHMATRIX: zero boxvolume   0.
 VOLNVT: move rejected, trial volume too low
 VOLNVT: a trial volume went negative
 should not happen in the NVT version of volume
Could anybody tell me what is the reason and how to fix it?
Thank you.
Li JG
Department of Chemical engineering
National University of Singapore



		
---------------------------------
Yahoo! for Good
 Click here to donate to the Hurricane Katrina relief effort. 

Dear all,
 
I want to define a new atom type to construct my molecule using UFF. I defined the atom type and its corresponding parameters in the forcefiled file: towhee_ff_UFF, but when runing the job, I got error message: parameters not found for atom 
Is there any other file should I modify or shoud I recompile it before running the job?
 
Thanks in advance,
 
Li JG
Department of chemical engineering
National University of Singapore



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

Dear Ozgur,
Thanks very much. when I complete that line with to 70 characters with space, the problem disappeared. I should read the manual carefully before asking .
Li JG
Department of chemical engineeirng
National University of Singapore

Ozgur YAZAYDIN <yaz...@WP...> wrote:
li jianguo wrote:
> Dear Yao£¬
> I did put the towhee_ff_UFF into the current directory and specify the
> directory in the input file, it is very strange that towhee could not find
> the forcefield file. Is there any other reason? Thanks a lot,
li jianguo wrote:
> Dear Yao£¬
> I did put the towhee_ff_UFF into the current directory and specify the
> directory in the input file, it is very strange that towhee could not find
> the forcefield file. Is there any other reason? Thanks a lot,

Dear Li,
I experienced a similar problem long time aga. It is strange, but in
some systems you may need to complete the input line where you specify
the location of the forcefield file to 70 characters. So just write the
full path of the forcefield file in your input and complete that line
with space characters to 70 characters. Actaully in towhee user manual
there is a part explaining this problem.

"On most systems you can just list this directory and then end the line.
However, if you have trouble then adding sufficient blank spaces to the
end of the line to get up to 70 characters could resolve your problem. "

I hope this helps.

-- 
Ozgur YAZAYDIN
Department of Chemical Engineering
Worcester Polytechnic Institute
100 Institute Rd.
Worcester, MA 01609-2280 USA




-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server.
Download it for free - -and be entered to win a 42" plasma tv or your very
own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php
_______________________________________________
Towhee-users mailing list
Tow...@li...
https://lists.sourceforge.net/lists/listinfo/towhee-users



		
---------------------------------
Yahoo! for Good
 Click here to donate to the Hurricane Katrina relief effort. 

li jianguo wrote:
> Dear Yao=A3=AC
> I did put the towhee_ff_UFF into the current directory and specify the
> directory in the input file, it is very strange that towhee could not f=
ind
> the forcefield file. Is there any other reason? Thanks a lot,
li jianguo wrote:
> Dear Yao=A3=AC
> I did put the towhee_ff_UFF into the current directory and specify the
> directory in the input file, it is very strange that towhee could not f=
ind
> the forcefield file. Is there any other reason? Thanks a lot,

Dear Li,
  I experienced a similar problem long time aga. It is strange, but in
some systems you may need to complete the input line where you specify
the location of the forcefield file to 70 characters. So just write the
full path of the forcefield file in your input and complete that line
with space characters to 70 characters. Actaully in towhee user manual
there is a part explaining this problem.

"On most systems you can just list this directory and then end the line.
However, if you have trouble then adding sufficient blank spaces to the
end of the line to get up to 70 characters could resolve your problem. "

I hope this helps.

--=20
Ozgur YAZAYDIN
Department of Chemical Engineering
Worcester Polytechnic Institute
100 Institute Rd.
Worcester, MA 01609-2280 USA

Dear Yao£¬
I did put the towhee_ff_UFF into the current directory and specify the directory in the input file, it is very strange that towhee could not find the forcefield file. Is there any other reason? Thanks a lot,
 
Li JG
Department of Chemical engineeirng
National University of Singapore

Yao Houndonougbo <ya...@ku...> wrote:
Dear Li,
If you get empty "towhee_ff_UFF" file, it means that the program did not find "towhee_ff_UFF" file as  specified in the towhhe_input file. You need to check the path of the towhee_ff_UFF file.

Yao.

li jianguo wrote: 
Dear all,

I met a problem when trying to simulate the phase diagram of some simle molecules using UFF, like H2.
I got the error message as follows:
...
pressurefreq:                   1
trmaxdispfreq:                  10
volmaxdispfreq:                  10
chempotperstep:         0
potentialstyle: classical                     
ffnumber:                   1
ff_filename:
                                                      
classical_potential: UFF 12-6                      
classical_mixrule: Geometric                     
cmix_rescaling_style: none                          
lshift: F
ltailc: T
rmin:   1.00000
rcut:  14.00000
rcutin:  10.00000
coulombstyle: none                
 Setting up force field parameters from files
 opening forcefield file:    1
 RWFORCEFIELD: problem with file number  1
 no data in ff_filename listed below:
                                      
when checking the files of the currrent directory, besides the forcefield file towhee_ff_UFF, there generated a new file: 'towhee_ff_UFF?' which is empty. Could anybody kindly clarify what is the possible reason and how to fix it? I also attached my input file with this email, thanks.

Li JG
Department of chemical engineering
National University of Singapore






---------------------------------
Yahoo! for Good
Click here to donate to the Hurricane Katrina relief effort. 

---------------------------------
randomseed1302002inputformat'Towhee'ensemble'nvt'temperature25.0d0nmolty1nmolectyp500numboxes1stepstyle'cycles'nstep1000printfreq 10   blocksize 50moviefreq100000backupfreq  10000runoutput'full'loutpdb.true.loutdft .false.  loutlammps.false.pressurefreq1trmaxdispfreq10volmaxdispfreq10chempotperstep0potentialstyle'classical'ffnumber1ff_filename'towhee_ff_UFF'classical_potential'UFF 12-6'classical_mixrule'Geometric'cmix_rescaling_style'none'lshift.false.ltailcTrmin  1.0d0 rcut  14.0d0rcutin 10.0d0 coulombstyle'none'nfield0isolvtype0linit   Tinitboxtype'dimensions'initstyle'coords'initlattice'none' initmol150350inix iniy iniz6    6    68    8    8hmatrix20.0d0 0.0d0 0.0d00.0d0 20.0d0 0.0d00.0d0 0.0d0 20.0d027.0d0 0.0d0 0.0d00.0d0 27.0d0 0.0d00.0d0 0.0d0 27.0d0pmvol0.1d0          pmvlpr          1.0d0          rmvol          20.0d0          tavol          0.5d0pmcell0.1d0          pmcellpr          1.0d0          pmcellpt          0.5d0          rmcell          2.0d0          tacell     
     0.5d0pm2boxrbswap1.0d0          pm2rbswmt          1.0d0          pm2rbswpr          1.0d0pm2boxcbswap1.0d0          pm2cbswmt          1.0d0          pm2cbswpr          1.0d0pm1boxcbswap0.2d0               pm1cbswmt          1.0d0pmavb10.0d0          pmavb1in          0.5d0          pmavb1mt          1.0d0          pmavb1ct          1.0d0          avb1rad          8.0d0pmavb20.0d0          pmavb2in          0.5d0          pmavb2mt          1.0d0          pmavb2ct          1.0d0          avb2rad          8.0d0pmavb30.0d0          pmavb3mt          1.0d0          pmavb3ct          1.0d0          avb3rad          8.0d0pmcb      0.5d0               pmcbmt          1.0d0          pmall          0.0d0 pmback0.0d0          pmbkmt          1.0d0pmpivot0.2d0          pmpivmt          1.0d0pmconrot0.0d0          pmcrmt	  1.0d0pmcrback0.0d0          pmcrbmt	  1.0d0pmplane0.0d0          pmplanebox          1.0d0          planewidth          3.0d0pmrow0.1d0          pmrowbox          1.0d0
 1.0d0          rowwidth          3.0d0pmtraat   1.0d0	            pmtamt          1.0d0          rmtraa          0.5d0           tatraa          0.5d0pmtracm   0.67d0	            pmtcmt          1.0d0          rmtrac          0.5d0           tatrac          0.5d0pmrotate1.0d0          pmromt            1.0d0          rmrot          0.05d0          tarot          0.5d0cbmc_style'coupled-decoupled'coupled_decoupled_form'Martin and Siepmann JPCB 1999'cbmc_setting_style'default ideal'inpstyle2nunit2nmaxcbmc2forcefield'UFF'charge_assignment'manual'unit ntype qqatom1    'H_'   0.0d0vibration12improper torsion0unit ntype qqatom2    'H_'   0.0d0vibration11improper torsion0  



-- -----------------------------------------------Yao Houndonougbo, Ph.D.Department of Chemistry and Center for Environmentally Beneficial Catalysis The University of Kansas1251 Wescoe Hall Drive, Room 2010 Malott HallLawrence, KS  66045-7582Tel:    (785) 864-3048Fax:    (785) 864-5396E-mail: ya...@ku...----------------------------------------------



		
---------------------------------
Yahoo! for Good
 Click here to donate to the Hurricane Katrina relief effort. 

Dear All,

I have some questions about DMD (Discontinuous Molecular Dynamics). Is 
there anyone who could help me?
AV
Cordially

Amir Vahid (ami...@iu...) 
Thermokinetic Research Laboratory
Chem. Eng. Dept.
Iran University of Science and Technology, Tehran, Iran 
Phone:+9821-73912731,fax: +9821-77607556
Postal Code: 16846-13114


-----Original Message-----
From: "Marcus Martin" <ma...@sa...>
To: tow...@li...
Cc: tow...@li...
Date: Sat, 17 Sep 2005 19:06:42 -0600
Subject: [Towhee-users] Re: [Towhee-developers] Torsion angle

> On Sep 16, 2005, at 7:01 AM, Laszlo Merenyi wrote:
> 
> > Could anybody kindly tell a tick, how to keep the torsion angles at a
> > fixed value?
> 
> Well, the default is to use one of the null torsions and then just 
> don't do any CBMC moves.  I don't think this helps you though.
> 
> > We have tried to use the simple harmonic torsion potential (Torsion 
> > style 1) with very high torcoeff(0) value. However, this solution 
> > arises some problem
> > in the uVT ensemble we use, as during particle insertion moves, the 
> > configuration
> > bias procedure has to try out many torsion angles to result in an 
> > acceptable structure.
> 
> Right.  And the reason there is no rigid torsion angle is nobody ever 
> wanted one before and it can lead to some confusing situation when you 
> start to
> have branched or cyclic molecules and are trying to do CBMC.  Plus, not
> that exciting coming up with moves to sample the conformation of a 
> rigid molecule.
> 
> The immediate solution would be to use a torsion that results in zero 
> energy, set up the system so your initial conformation is the one you 
> desire, and then
> don't do any of the molecule sampling moves.  This will not work with 
> the grand canonical ensemble in Towhee, but you could perform a two box
> NpT Gibbs ensemble where you only perform volume moves in box 2, and 
> use the rotational-bias 2 box swap move.  This results in a simulation 
> that is extremely similar
> to the Grand Canonical ensemble, but with T and p specified instead of 
> chemical potential.  Works fine so long as you are only interested in 
> performing grand canonical moves on a single component of the system.
> 
> If that doesn't work then send a detailed email to the 
> towhee-developers list with your towhee_input and initial coordinates 
> file and I'll see if there is a not too painful way to add rigid 
> torsions to the code to make this work for your molecule.
> 
> > We suppose that it is a small error in the code, that when
> > torsion style is 1 , the program requires three coefficients
> > in the forcefield-file, however only two (torcoeff(0) and
> torcoeff(1))
> > describe the potential. See lines 1209-1215 in the rwforcefiled.F 
> > source file,
> > where either coeffstart or coeffstop should be set to 1.
> 
> Looks like a bug to me.  Fixing that for version 4.11.4.  Also making 
> some more bug fixes for the new CBMC moves and some other changes that 
> will affect the measured chemical potential for molecules with flexible
> bonds.  More on that next week.
> 
> Marcus
> -------------------------------------------------------------------
> Marcus G. Martin
> Senior Member of Technical Staff
> Multiscale Computational Materials Methods
> Sandia National Laboratories
> PO Box 5800 MS 1110
> Albuquerque NM 87185-1110
> v. 505 284-6355
> f. 505 845-7442
> http://www.cs.sandia.gov/marcusmartin/
> http://towhee.sourceforge.net
> 
> 
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by:
> Tame your development challenges with Apache's Geronimo App Server. 
> Download it for free - -and be entered to win a 42" plasma tv or your
> very
> own Sony(tm)PSP.  Click here to play:
> http://sourceforge.net/geronimo.php
> _______________________________________________
> Towhee-users mailing list
> Tow...@li...
> https://lists.sourceforge.net/lists/listinfo/towhee-users

Dear Li,
If you get empty "towhee_ff_UFF" file, it means that the program did not 
find "towhee_ff_UFF" file as  specified in the towhhe_input file. You 
need to check the path of the towhee_ff_UFF file.

Yao.

li jianguo wrote:

> Dear all,
>
> I met a problem when trying to simulate the phase diagram of some 
> simle molecules using UFF, like H2.
> I got the error message as follows:
> ...
> pressurefreq:                   1
> trmaxdispfreq:                  10
> volmaxdispfreq:                  10
> chempotperstep:         0
> potentialstyle: classical                    
> ffnumber:                   1
> ff_filename:
>                                                      
> classical_potential: UFF 12-6                     
> classical_mixrule: Geometric                    
> cmix_rescaling_style: none                         
> lshift: F
> ltailc: T
> rmin:   1.00000
> rcut:  14.00000
> rcutin:  10.00000
> coulombstyle: none               
>  Setting up force field parameters from files
>  opening forcefield file:    1
>  RWFORCEFIELD: problem with file number  1
>  no data in ff_filename listed below:
>                                      
> when checking the files of the currrent directory, besides the 
> forcefield file towhee_ff_UFF, there generated a new file: 
> 'towhee_ff_UFF?' which is empty. Could anybody kindly clarify what is 
> the possible reason and how to fix it? I also attached my input file 
> with this email, thanks.
>
> Li JG
> Department of chemical engineering
> National University of Singapore
>
>
>
> ------------------------------------------------------------------------
> Yahoo! for Good
> Click here to donate <http://store.yahoo.com/redcross-donate3/> to the 
> Hurricane Katrina relief effort.
>
>------------------------------------------------------------------------
>
>randomseed
>1302002
>inputformat
>'Towhee'
>ensemble
>'nvt'
>temperature
>25.0d0
>nmolty
>1
>nmolectyp
>500
>numboxes
>1
>stepstyle
>'cycles'
>nstep
>1000
>printfreq 
>10   
>blocksize 
>50
>moviefreq
>100000
>backupfreq  
>10000
>runoutput
>'full'
>loutpdb
>.true.
>loutdft 
>.false.  
>loutlammps
>.false.
>pressurefreq
>1
>trmaxdispfreq
>10
>volmaxdispfreq
>10
>chempotperstep
>0
>potentialstyle
>'classical'
>ffnumber
>1
>ff_filename
>'towhee_ff_UFF'
>classical_potential
>'UFF 12-6'
>classical_mixrule
>'Geometric'
>cmix_rescaling_style
>'none'
>lshift
>.false.
>ltailc
>T
>rmin  
>1.0d0 
>rcut  
>14.0d0
>rcutin 
>10.0d0 
>coulombstyle
>'none'
>nfield
>0
>isolvtype
>0
>linit   
>T
>initboxtype
>'dimensions'
>initstyle
>'coords'
>initlattice
>'none' 
>initmol
>150
>350
>inix iniy iniz
>6    6    6
>8    8    8
>hmatrix
>20.0d0 0.0d0 0.0d0
>0.0d0 20.0d0 0.0d0
>0.0d0 0.0d0 20.0d0
>27.0d0 0.0d0 0.0d0
>0.0d0 27.0d0 0.0d0
>0.0d0 0.0d0 27.0d0
>pmvol
>0.1d0
>          pmvlpr
>          1.0d0
>          rmvol
>          20.0d0
>          tavol
>          0.5d0
>pmcell
>0.1d0
>          pmcellpr
>          1.0d0
>          pmcellpt
>          0.5d0
>          rmcell
>          2.0d0
>          tacell
>          0.5d0
>pm2boxrbswap
>1.0d0
>          pm2rbswmt
>          1.0d0
>          pm2rbswpr
>          1.0d0
>pm2boxcbswap
>1.0d0
>          pm2cbswmt
>          1.0d0
>          pm2cbswpr
>          1.0d0
>
>pm1boxcbswap
>0.2d0     
>          pm1cbswmt
>          1.0d0
>pmavb1
>0.0d0
>          pmavb1in
>          0.5d0
>          pmavb1mt
>          1.0d0
>          pmavb1ct
>          1.0d0
>          avb1rad
>          8.0d0
>pmavb2
>0.0d0
>          pmavb2in
>          0.5d0
>          pmavb2mt
>          1.0d0
>          pmavb2ct
>          1.0d0
>          avb2rad
>          8.0d0
>pmavb3
>0.0d0
>          pmavb3mt
>          1.0d0
>          pmavb3ct
>          1.0d0
>          avb3rad
>          8.0d0
>pmcb      
>0.5d0     
>          pmcbmt
>          1.0d0
>          pmall
>          0.0d0 
>pmback
>0.0d0
>          pmbkmt
>          1.0d0
>pmpivot
>0.2d0
>          pmpivmt
>          1.0d0
>pmconrot
>0.0d0
>          pmcrmt
>	  1.0d0
>pmcrback
>0.0d0
>          pmcrbmt
>	  1.0d0
>pmplane
>0.0d0
>          pmplanebox
>          1.0d0
>          planewidth
>          3.0d0
>pmrow
>0.1d0
>          pmrowbox
>          1.0d0 1.0d0
>          rowwidth
>          3.0d0
>pmtraat   
>1.0d0	  
>          pmtamt
>          1.0d0
>          rmtraa
>          0.5d0 
>          tatraa
>          0.5d0
>pmtracm   
>0.67d0	  
>          pmtcmt
>          1.0d0
>          rmtrac
>          0.5d0 
>          tatrac
>          0.5d0
>pmrotate
>1.0d0
>          pmromt  
>          1.0d0
>          rmrot
>          0.05d0
>          tarot
>          0.5d0
>cbmc_style
>'coupled-decoupled'
>coupled_decoupled_form
>'Martin and Siepmann JPCB 1999'
>cbmc_setting_style
>'default ideal'
>inpstyle
>2
>nunit
>2
>nmaxcbmc
>2
>forcefield
>'UFF'
>charge_assignment
>'manual'
>unit ntype qqatom
>1    'H_'   0.0d0
>vibration
>1
>2
>improper torsion
>0
>unit ntype qqatom
>2    'H_'   0.0d0
>vibration
>1
>1
>improper torsion
>0
>  
>


-- 
-----------------------------------------------
Yao Houndonougbo, Ph.D.
Department of Chemistry and 
Center for Environmentally Beneficial Catalysis 
The University of Kansas
1251 Wescoe Hall Drive, Room 2010 Malott Hall
Lawrence, KS  66045-7582
Tel:    (785) 864-3048
Fax:    (785) 864-5396
E-mail: ya...@ku...
----------------------------------------------

randomseed
1302002
inputformat
'Towhee'
ensemble
'nvt'
temperature
25.0d0
nmolty
1
nmolectyp
500
numboxes
1
stepstyle
'cycles'
nstep
1000
printfreq 
10   
blocksize 
50
moviefreq
100000
backupfreq  
10000
runoutput
'full'
loutpdb
.true.
loutdft 
.false.  
loutlammps
.false.
pressurefreq
1
trmaxdispfreq
10
volmaxdispfreq
10
chempotperstep
0
potentialstyle
'classical'
ffnumber
1
ff_filename
'towhee_ff_UFF'
classical_potential
'UFF 12-6'
classical_mixrule
'Geometric'
cmix_rescaling_style
'none'
lshift
.false.
ltailc
T
rmin  
1.0d0 
rcut  
14.0d0
rcutin 
10.0d0 
coulombstyle
'none'
nfield
0
isolvtype
0
linit   
T
initboxtype
'dimensions'
initstyle
'coords'
initlattice
'none' 
initmol
150
350
inix iniy iniz
6    6    6
8    8    8
hmatrix
20.0d0 0.0d0 0.0d0
0.0d0 20.0d0 0.0d0
0.0d0 0.0d0 20.0d0
27.0d0 0.0d0 0.0d0
0.0d0 27.0d0 0.0d0
0.0d0 0.0d0 27.0d0
pmvol
0.1d0
          pmvlpr
          1.0d0
          rmvol
          20.0d0
          tavol
          0.5d0
pmcell
0.1d0
          pmcellpr
          1.0d0
          pmcellpt
          0.5d0
          rmcell
          2.0d0
          tacell
          0.5d0
pm2boxrbswap
1.0d0
          pm2rbswmt
          1.0d0
          pm2rbswpr
          1.0d0
pm2boxcbswap
1.0d0
          pm2cbswmt
          1.0d0
          pm2cbswpr
          1.0d0

pm1boxcbswap
0.2d0     
          pm1cbswmt
          1.0d0
pmavb1
0.0d0
          pmavb1in
          0.5d0
          pmavb1mt
          1.0d0
          pmavb1ct
          1.0d0
          avb1rad
          8.0d0
pmavb2
0.0d0
          pmavb2in
          0.5d0
          pmavb2mt
          1.0d0
          pmavb2ct
          1.0d0
          avb2rad
          8.0d0
pmavb3
0.0d0
          pmavb3mt
          1.0d0
          pmavb3ct
          1.0d0
          avb3rad
          8.0d0
pmcb      
0.5d0     
          pmcbmt
          1.0d0
          pmall
          0.0d0 
pmback
0.0d0
          pmbkmt
          1.0d0
pmpivot
0.2d0
          pmpivmt
          1.0d0
pmconrot
0.0d0
          pmcrmt
	  1.0d0
pmcrback
0.0d0
          pmcrbmt
	  1.0d0
pmplane
0.0d0
          pmplanebox
          1.0d0
          planewidth
          3.0d0
pmrow
0.1d0
          pmrowbox
          1.0d0 1.0d0
          rowwidth
          3.0d0
pmtraat   
1.0d0	  
          pmtamt
          1.0d0
          rmtraa
          0.5d0 
          tatraa
          0.5d0
pmtracm   
0.67d0	  
          pmtcmt
          1.0d0
          rmtrac
          0.5d0 
          tatrac
          0.5d0
pmrotate
1.0d0
          pmromt  
          1.0d0
          rmrot
          0.05d0
          tarot
          0.5d0
cbmc_style
'coupled-decoupled'
coupled_decoupled_form
'Martin and Siepmann JPCB 1999'
cbmc_setting_style
'default ideal'
inpstyle
2
nunit
2
nmaxcbmc
2
forcefield
'UFF'
charge_assignment
'manual'
unit ntype qqatom
1    'H_'   0.0d0
vibration
1
2
improper torsion
0
unit ntype qqatom
2    'H_'   0.0d0
vibration
1
1
improper torsion
0

On Sep 16, 2005, at 7:01 AM, Laszlo Merenyi wrote:

> Could anybody kindly tell a tick, how to keep the torsion angles at a 
> fixed value?

Well, the default is to use one of the null torsions and then just 
don't do any CBMC moves.  I don't think this helps you though.

> We have tried to use the simple harmonic torsion potential (Torsion 
> style 1) with very high torcoeff(0) value. However, this solution 
> arises some problem
> in the uVT ensemble we use, as during particle insertion moves, the 
> configuration
> bias procedure has to try out many torsion angles to result in an 
> acceptable structure.

Right.  And the reason there is no rigid torsion angle is nobody ever 
wanted one before and it can lead to some confusing situation when you 
start to
have branched or cyclic molecules and are trying to do CBMC.  Plus, not 
that exciting coming up with moves to sample the conformation of a 
rigid molecule.

The immediate solution would be to use a torsion that results in zero 
energy, set up the system so your initial conformation is the one you 
desire, and then
don't do any of the molecule sampling moves.  This will not work with 
the grand canonical ensemble in Towhee, but you could perform a two box 
NpT Gibbs ensemble where you only perform volume moves in box 2, and 
use the rotational-bias 2 box swap move.  This results in a simulation 
that is extremely similar
to the Grand Canonical ensemble, but with T and p specified instead of 
chemical potential.  Works fine so long as you are only interested in 
performing grand canonical moves on a single component of the system.

If that doesn't work then send a detailed email to the 
towhee-developers list with your towhee_input and initial coordinates 
file and I'll see if there is a not too painful way to add rigid 
torsions to the code to make this work for your molecule.

> We suppose that it is a small error in the code, that when
> torsion style is 1 , the program requires three coefficients
> in the forcefield-file, however only two (torcoeff(0) and torcoeff(1))
> describe the potential. See lines 1209-1215 in the rwforcefiled.F 
> source file,
> where either coeffstart or coeffstop should be set to 1.

Looks like a bug to me.  Fixing that for version 4.11.4.  Also making 
some more bug fixes for the new CBMC moves and some other changes that 
will affect the measured chemical potential for molecules with flexible 
bonds.  More on that next week.

Marcus
-------------------------------------------------------------------
Marcus G. Martin
Senior Member of Technical Staff
Multiscale Computational Materials Methods
Sandia National Laboratories
PO Box 5800 MS 1110
Albuquerque NM 87185-1110
v. 505 284-6355
f. 505 845-7442
http://www.cs.sandia.gov/marcusmartin/
http://towhee.sourceforge.net

Dear Laszlo,
You just need to set all the torsion parameters to 0.0. For example you 
can use the torsion style 13 and set the two coefficients to 0.0 and  
the dihedral energy should be zero.

Yao.

Laszlo Merenyi wrote:

> Dear All,
>
> We want to run Towhee simulations using multi-centre rigid
> molecule model. We define the molecule structure in towhee_template.
> As the molecule is rigid, we define fix bond lengths (Bond style 1)
> and rigid angles (Angle style 0) in the forcefield file. When we start
> the simulation, Towhee requests torsion terms, as well. According to 
> the manual,
> there is no fixed torsion style.
> Could anybody kindly tell a tick, how to keep the torsion angles at a 
> fixed value?
> We have tried to use the simple harmonic torsion potential (Torsion 
> style 1) with very high torcoeff(0) value. However, this solution 
> arises some problem
> in the uVT ensemble we use, as during particle insertion moves, the 
> configuration
> bias procedure has to try out many torsion angles to result in an 
> acceptable structure.
>
> We suppose that it is a small error in the code, that when
> torsion style is 1 , the program requires three coefficients
> in the forcefield-file, however only two (torcoeff(0) and torcoeff(1))
> describe the potential. See lines 1209-1215 in the rwforcefiled.F 
> source file,
> where either coeffstart or coeffstop should be set to 1.
>
>
> Thank for your advice in advance,
>
> Laszlo Merenyi
>
> Eotvos Lorand Geophysical Institute of Hungary
>
>
> -------------------------------------------------------
> SF.Net email is sponsored by:
> Tame your development challenges with Apache's Geronimo App Server. 
> Download it for free - -and be entered to win a 42" plasma tv or your 
> very
> own Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
> _______________________________________________
> Towhee-users mailing list
> Tow...@li...
> https://lists.sourceforge.net/lists/listinfo/towhee-users
>


-- 
-----------------------------------------------
Yao Houndonougbo, Ph.D.
Department of Chemistry and 
Center for Environmentally Beneficial Catalysis 
The University of Kansas
1251 Wescoe Hall Drive, Room 2010 Malott Hall
Lawrence, KS  66045-7582
Tel:    (785) 864-3048
Fax:    (785) 864-5396
E-mail: ya...@ku...
----------------------------------------------

Dear All,

We want to run Towhee simulations using multi-centre rigid
molecule model. We define the molecule structure in towhee_template.
As the molecule is rigid, we define fix bond lengths (Bond style 1)
and rigid angles (Angle style 0) in the forcefield file. When we start
the simulation, Towhee requests torsion terms, as well. According to the 
manual,
there is no fixed torsion style.
Could anybody kindly tell a tick, how to keep the torsion angles at a 
fixed value?
We have tried to use the simple harmonic torsion potential (Torsion 
style 1) 
with very high torcoeff(0) value. However, this solution arises some problem
in the uVT ensemble we use, as during particle insertion moves, the 
configuration
bias procedure has to try out many torsion angles to result in an 
acceptable structure.

We suppose that it is a small error in the code, that when
torsion style is 1 , the program requires three coefficients
in the forcefield-file, however only two (torcoeff(0) and torcoeff(1))
describe the potential. See lines 1209-1215 in the rwforcefiled.F source 
file,
where either coeffstart or coeffstop should be set to 1.


Thank for your advice in advance,

Laszlo Merenyi

Eotvos Lorand Geophysical Institute of Hungary

On Sep 14, 2005, at 1:55 AM, li jianguo wrote:

>  Now I am working on simulation of non-shperical macromolecules like=20=

> protein or=A0 polymers. I will use simplified models, for example, I=20=

> will simplify the polymer into a three-site or a four site model, each=20=

> site is treated as a sphere. If we define each sphere as a fake atom=20=

> and add connections between some spheres, then we get a fake molecule=20=

> which has simple structure. Is it possible to define such fake=20
> molecules and simulate its phase diagram by Towhee? Any suggestions?

You can create any force field you desire and use it in Towhee.  Simply=20=

construct the appropriate towhee_ff_yourforcefieldhere file with the=20
interaction sites
having the parameters you desire and the bonds being something=20
appropriate for your coarse grain model.

The format of the towhee_ff files is found in the Towhee web manual

http://towhee.sourceforge.net/towhee_ff.html

I personally prefer to create a subroutine full of the parameters and=20
then use the built in utilities in Towhee to create the very particular=20=

format required by the code.
This uses the 'forcefield' utility that can be constructed in the=20
Source directory.  Directions are available at

http://towhee.sourceforge.net/code/code_manual.html

If you are taking that route you would want to construct an=20
ffyourforcefieldhere.F subroutine (based upon one of the many already=20
in the code) and then modify
ffcreate.F in order to incorporate your new forcefield into the=20
forcefield builder utility.

As for your question about the meaning of the variables, Matt is=20
correct that many of the variables are described in the users manual as=20=

they tend to be input parameters.  There is no other comprehensive list=20=

of variable meanings as no one has felt motivated to construct such a=20
list.  The globalf.h file contains all of the global fortran variables=20=

and sometimes has a useful comment about what they mean.

Marcus

-------------------------------------------------------------------
Marcus G. Martin
Senior Member of Technical Staff
Multiscale Computational Materials Methods
Sandia National Laboratories
PO Box 5800 MS 1110
Albuquerque NM 87185-1110
v. 505 284-6355
f. 505 845-7442
http://www.cs.sandia.gov/marcusmartin/
http://towhee.sourceforge.net

Dear all,

I want to simulate the phase diagram of macromolecules like polymers or proteins. To save the computation time, I will divide the macromolecules into several parts, each parts is treated as one sphere with a L-J potential, thus the macromolecules can be seen as a multi-site model, though each site contains many atoms, thus the multi-site model is just like a fake molecule. This model can roughly reflect the shape of the macromolecule. We add connections between the spheres within each molecule and do not allow intramolecular interaction, then the structure of the fake molecule is fixed. The intermolecular interaction is composed of site-site interaction between different molecules. I am wondering is it possible to construct such a fake molecule in Towhee? And any suggestions of implementing such a simulation?

Thanks a lot.

best regards,

Li JG

Department of chemical engineering

National University of Singapore

 




		
---------------------------------
Yahoo! for Good
 Click here to donate to the Hurricane Katrina relief effort. 

Many of the variables have the same name and meaning as they do in, e.g. 
the towhee_input file.  They typically have logical names (logical 
variables begin with 'l', integers with 'n').  You can see if a variable 
is global by checking to see if it is defined in 'globalf.h'.  
Otherwise, the only way discover the meaning of a variable is to read 
through the code: see where the variables are set, how they are used, 
etc.  Doing a search for a variable name across all the files (on UNIX, 
'grep string *.F') is very helpful. 

If you are stuck, feel free to write to the mailing list (Marcus: is 
this discussion better suited for the towhee-developers list?).  
Finally, as you do this, consider writing up what you learned and 
posting it to the list: such notes could form the basis of very helpful 
documentation.

Best,

Matt

li jianguo wrote:

> Dear all,
>  
> I am reading the codes  and try to understand them. The problem is 
> that it is difficult to guess the meaning of the variables in the 
> codes, anybody have a list of the meaning of the variables appeared in 
> the codes? Thanks a lot.


-- 
Matt Wyczalkowski
Graduate Student, Biomedical Engineering
Pappu Lab: http://lima.wustl.edu
Washington University in St. Louis
ma...@ce...

Dear all,
 
I am reading the codes  and try to understand them. The problem is that it is difficult to guess the meaning of the variables in the codes, anybody have a list of the meaning of the variables appeared in the codes? Thanks a lot.
 
Li JG
Department of chemical engineering
National University of Singapore  



		
---------------------------------
Yahoo! for Good
 Click here to donate to the Hurricane Katrina relief effort. 

Dear all£¬
First thank Wyczalkowski for your suggestion. 
Now I am working on simulation of non-shperical macromolecules like protein or  polymers. I will use simplified models, for example, I will simplify the polymer into a three-site or a four site model, each site is treated as a sphere. If we define each sphere as a fake atom and add connections between some spheres, then we get a fake molecule which has simple structure. Is it possible to define such fake molecules and simulate its phase diagram by Towhee? Any suggestions?
Thanks again,
best regards,
Li JG
Department of Chemical engineering
National University of Singapore

Matt Wyczalkowski <ma...@ce...> wrote:
I just implemented a new potential, 'Scaled Lennard-Jones', released in 
the 4.11.0 version. Let me briefly tell you what I learned, and I'm 
sure Marcus will be able to fill in the details and correct the mistakes.

The potential energy is calculated in one of two ways during the 
execution of the program: it can be calculated all at once with the 
engtotal() function, or it can be calculated incrementally, at each 
move. For the most part, the latter method is used, as it is faster. 
See, for example, line 228 in mainloop.F, where a call to engtotal sets 
the initial energies, and the array vctotal (which holds total system 
energy), and others are initialized; thereafter, these arrays are 
incremented at each move (e.g. line 172 in rotate.F). A final call to 
engtotal (line 1620 in mainloop.F) does a sanity check, confirming that 
incrementally updated energies match the globally calculated energy.

That said, you probably don't have to modify engtotal or any of the 
incremental updates. You simply need to modify the functions where the 
potentials are calculated. The function vtwobody() calculates the 
two-body potential, and wtwobody() does related two-body calculations to 
obtain the pressure. If you will be implementing analytical cutoff 
corrections, you will need to modify tail.F.

These are the main files needed to implement the potential, though 
you'll also have to make minor changes in the input logic and other 
places. You might want to do a global search for the string 
'classical_potential' in the code, to see where the implementations lie.

It would also be very helpful to look at the changes I made in going 
from version 4.10.x to 4.11.0; see 
http://cvs.sourceforge.net/viewcvs.py/towhee/towhee/ . This set of 
changes contains just about everything needed to implement the new 
'Scaled Lennard-Jones' potential.

Good luck!

Matt

li jianguo wrote:

> Dear all,
> 
> I want to simulate colloid phase transition using the some type of 
> potential which is not included in the towhee forcefield, therefore I 
> should modify the code to add the my potential, I searched the code 
> but did not find the subroutine for calculating the potential energy. 
> Could anybody tell me which file should I modify and how to modify it 
> in order to use other potentials?
> Thanks in advance,
> 
> Li JG
> Department of Chemical engineering
> National University of Singapore 
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>


-- 
Matt Wyczalkowski
Graduate Student, Biomedical Engineering
Pappu Lab: http://lima.wustl.edu
Washington University in St. Louis
ma...@ce...




-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
Towhee-users mailing list
Tow...@li...
https://lists.sourceforge.net/lists/listinfo/towhee-users



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

I just implemented a new potential, 'Scaled Lennard-Jones', released in 
the 4.11.0 version.  Let me briefly tell you what I learned, and I'm 
sure Marcus will be able to fill in the details and correct the mistakes.

The potential energy is calculated in one of two ways during the 
execution of the program: it can be calculated all at once with the 
engtotal() function, or it can be calculated incrementally, at each 
move.  For the most part, the latter method is used, as it is faster.  
See, for example, line 228 in mainloop.F, where a call to engtotal sets 
the initial energies, and the array vctotal (which holds total system 
energy), and others are initialized; thereafter, these arrays are 
incremented at each move (e.g. line 172 in rotate.F).  A final call to 
engtotal (line 1620 in mainloop.F) does a sanity check, confirming that 
incrementally updated energies match the globally calculated energy.

That said, you probably don't have to modify engtotal or any of the 
incremental updates.  You simply need to modify the functions where the 
potentials are calculated.  The function vtwobody() calculates the 
two-body potential, and wtwobody() does related two-body calculations to 
obtain the pressure.  If you will be implementing analytical cutoff 
corrections, you will need to modify tail.F.

These are the main files needed to implement the potential, though 
you'll also have to make minor changes in the input logic and other 
places.  You might want to do a global search for the string 
'classical_potential' in the code, to see where the implementations lie.

It would also be very helpful to look at the changes I made in going 
from version 4.10.x to 4.11.0; see 
http://cvs.sourceforge.net/viewcvs.py/towhee/towhee/ .  This set of 
changes contains just about everything needed to implement the new 
'Scaled Lennard-Jones' potential.

Good luck!

Matt

li jianguo wrote:

> Dear all,
>  
> I want to simulate colloid phase transition using the some type of 
> potential which is not included in the towhee forcefield, therefore I 
> should modify the code to add the my potential, I searched the code 
> but did not find the subroutine for calculating the potential energy. 
> Could anybody tell me which file should I modify and how to modify it 
> in order to use other potentials?
> Thanks in advance,
>  
> Li JG
> Department of Chemical engineering
> National University of Singapore  
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>


-- 
Matt Wyczalkowski
Graduate Student, Biomedical Engineering
Pappu Lab: http://lima.wustl.edu
Washington University in St. Louis
ma...@ce...

Dear all,
 
I want to simulate colloid phase transition using the some type of potential which is not included in the towhee forcefield, therefore I should modify the code to add the my potential, I searched the code but did not find the subroutine for calculating the potential energy. Could anybody tell me which file should I modify and how to modify it in order to use other potentials? 
Thanks in advance,
 
Li JG
Department of Chemical engineering
National University of Singapore  



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

On Sep 5, 2005, at 9:15 AM, li jianguo wrote:

> I met a problem when calculating the vaper-liquid phase diagram=A0using=20=

> Lennar-Jones potential. In the end of the towhee.out, I found that the=20=

> virial pressures=A0of the two boxes don't equal=A0 and also have=A0 =
large=20
> fluctuations even if the density of each box has been=A0almost =
constant.=20
> I tried to do simulaiton with more cycles, but the virial pressures of=20=

> the two boxes=A0never equal. I wonder whether the system=A0can be=20
> considered=A0in equilibrium when the virial pressure of the two boxes=20=

> don't equal. Anybody can kindly tell what may be the reason and how to=20=

> fix this problem?

I would need a lot more specific information about the system to decide=20=

whether this was a problem.  Send a report with the input files to
towhee-bugs

Otherwise the virial pressures should be equal if the two phases are in=20=

equilibrium.  However, the liquid box pressure tends to have error bars=20=

on the order of 1000 kPa so you can get strange looking averages (even=20=

negative pressures) just because it is so noisy.

Marcus
-------------------------------------------------------------------
Marcus G. Martin
Senior Member of Technical Staff
Multiscale Computational Materials Methods
Sandia National Laboratories
PO Box 5800 MS 1110
Albuquerque NM 87185-1110
v. 505 284-6355
f. 505 845-7442
http://www.cs.sandia.gov/marcusmartin/
http://towhee.sourceforge.net