IBI for bonded

[Valentina]

Hello!

I am a little confused here.

I have read somewhere that Votca 1.3 (? I thought latest version is 1.2.3) supports IBI for bonded - is this right?

If so, how do I implement it? I have attatched my settings file (saved in .odt,but used as .xml on the machine). This doesn't seem to do more then fitting the nonbonded C-C potential. Am I missing something?

Many thanks!
Valentina

[Christoph Junghans]

Hi,

2012/9/28 Valentina <valentina...@yahoo.co.uk>:

> Hello!
>
> I am a little confused here.
>
> I have read somewhere that Votca 1.3 (? I thought latest version is 1.2.3)
> supports IBI for bonded - is this right?

Yes, it is implemented in the devlopment version and will be released in 1.3.
You can build it using the "--dev" option of build.sh,

>
> If so, how do I implement it? I have attatched my settings file (saved in
> .odt,but used as .xml on the machine). This doesn't seem to do more then
> fitting the nonbonded C-C potential. Am I missing something?

There is an example in the tutorials:
<http://code.google.com/p/votca/source/browse/?repo=csg-tutorials#hg%2Fhexane%2Fibi_bonded>

Your settings file looks good, except you will need an xml mapping
file (cg.inverse.map) to make VOTCA understand. which atoms belong to
which bond and angle.
Have a look at hexane_cg.xml in above example. The best is to run
csg_stat with "--cg map.xml" option by hand first to see if the
distributions are right.
And there is one more glitch, every bonded interaction will need a
potential input (pot.in file).

Cheers,

Christoph

>
> Many thanks!
> Valentina
>
> --
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/votca/-/YpupeBEK0_4J.
> To post to this group, send email to vo...@googlegroups.com.
> To unsubscribe from this group, send email to
> votca+un...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/votca?hl=en.



--
Christoph Junghans
Web: http://www.compphys.de

[Valentina]

Thank you very much, Christoph!

I already have all of the potentials:) I was just little confused with versions and which one does IBI for bonded; and if I am not specifying things correctly in the settigs.xml

I think I have 1.2.3 installed atm, so I am refining non-bonded for this system. 

I have done IBI bonded with our own code already;) but I was hoping I can do it within Votca.

Have a good weekend,

Valentina

[Valentina]

Hello!

I am trying to install votca with --dev and getting this:

done with csg
Unknown progamm '--dev', I know: tools csg csg-tutorials csgapps csg-testsuite csg-manual gromacs (disable this check with --no-progcheck option)

How do I get the dev version? I need to do IBI for bonded.

Thanks!

[Christoph Junghans]

2012/10/16 Valentina <valentina...@yahoo.co.uk>:
> Hello!
>

> I am trying to install votca with --dev and getting this:
>
> done with csg
> Unknown progamm '--dev', I know: tools csg csg-tutorials csgapps
> csg-testsuite csg-manual gromacs (disable this check with --no-progcheck
> option)

I guess, you mixed up the order, "--dev" has to come before the program names:
$ ./build.sh --prefix ${prefix} --dev tools csg

Christoph

> https://groups.google.com/d/msg/votca/-/56wJuKUTv7kJ.

[Valentina]

Yes, you are absolutely right. I forgot there is meant to be some order. Thank you!

[Valentina]

I am now trying to install th developer version on the cluster:

./build.sh -DWITH_SQLITE3=NO -DEXTERNAL_BOOST=OFF -DGSL_INCLUDE_DIR=/share/apps/gsl/include -DGSL_LIBRARY=/share/apps/gsl/lib/libgsl.so -DGSLCBLAS_LIBRARY=/share/apps/gsl/lib/libgslcblas.so -DFFTW3_INCLUDE_DIR=/share/apps/fftw3_d/include -DFFTW3_LIBRARY=/share/apps/fftw3_d/lib/libfftw3.so -DGROMACS_INCLUDE_DIR=/share/apps/gromacs/include -DGROMACS_LIBRARY=/share/apps/gromacs/lib/libgmx.so -DGROMACS_DEP_LIBARIES="/share/apps/gsl/lib/libgslcblas.so;/share/apps/gsl/lib/libgsl.so" --prefix /share/apps/votca --dev tools csg

This line worked for the last installation we have done
(I have also tried-DEXTERNAL_BOOST=NO )

and I am getting the error. Could you advise me, please, where I am going wrong

Thank you

CMake Error at CMakeModules/FindBoost.cmake:912 (message):
  Unable to find the requested Boost libraries.

  Unable to find the Boost header files.  Please set BOOST_ROOT to the root
  directory containing Boost or BOOST_INCLUDEDIR to the directory containing
  Boost's headers.
Call Stack (most recent call first):
  src/CMakeLists.txt:1 (find_package)


-- Found EXPAT: /usr/lib64/libexpat.so 
-- checking for module 'fftw3'
--   package 'fftw3' not found
-- Found FFTW3: /share/apps/fftw3_d/lib/libfftw3.so 
-- Looking for fftw_plan_r2r_1d in /share/apps/fftw3_d/lib/libfftw3.so;m
-- Looking for fftw_plan_r2r_1d in /share/apps/fftw3_d/lib/libfftw3.so;m - found
-- checking for module 'gsl'
--   package 'gsl' not found
-- Found GSL: /share/apps/gsl/lib/libgsl.so 
-- Looking for cblas_dsyrk in /share/apps/gsl/lib/libgslcblas.so;m
-- Looking for cblas_dsyrk in /share/apps/gsl/lib/libgslcblas.so;m - found
-- Looking for gsl_linalg_QR_decomp in /share/apps/gsl/lib/libgsl.so;/share/apps/gsl/lib/libgslcblas.so;m
-- Looking for gsl_linalg_QR_decomp in /share/apps/gsl/lib/libgsl.so;/share/apps/gsl/lib/libgslcblas.so;m - found
-- Performing Test HAVE_AIX
-- Performing Test HAVE_AIX - Failed
-- Could NOT find Doxygen (missing:  DOXYGEN_EXECUTABLE)
CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
Boost_INCLUDE_DIR (ADVANCED)
   used as include directory in directory /root/tools/src/libtools
   used as include directory in directory /root/tools/src/libtools
   used as include directory in directory /root/tools/src/tools
Boost_PROGRAM_OPTIONS_LIBRARY (ADVANCED)
    linked by target "votca_tools" in directory /root/tools/src/libtools

-- Configuring incomplete, errors occurred!

[Christoph Junghans]

2012/10/16 Valentina <valentina...@yahoo.co.uk>:

> I am now trying to install th developer version on the cluster:
>
> ./build.sh -DWITH_SQLITE3=NO -DEXTERNAL_BOOST=OFF
> -DGSL_INCLUDE_DIR=/share/apps/gsl/include
> -DGSL_LIBRARY=/share/apps/gsl/lib/libgsl.so
> -DGSLCBLAS_LIBRARY=/share/apps/gsl/lib/libgslcblas.so
> -DFFTW3_INCLUDE_DIR=/share/apps/fftw3_d/include
> -DFFTW3_LIBRARY=/share/apps/fftw3_d/lib/libfftw3.so
> -DGROMACS_INCLUDE_DIR=/share/apps/gromacs/include
> -DGROMACS_LIBRARY=/share/apps/gromacs/lib/libgmx.so
> -DGROMACS_DEP_LIBARIES="/share/apps/gsl/lib/libgslcblas.so;/share/apps/gsl/lib/libgsl.so"
> --prefix /share/apps/votca --dev tools csg
>
> This line worked for the last installation we have done
> (I have also tried-DEXTERNAL_BOOST=NO )

Due to a lack of manpower and other reason, we have removed the
bundled version of boost in VOTCA 1.3 a while ago (see my other
email), that is why boost is now a mandatory dependency.

I am sorry for that, but you will have to install it or get it
installed by your admin.

Christoph

> https://groups.google.com/d/msg/votca/-/rqongRKdI28J.

[Valentina]

Ok, I will install it (and update relevant installation manuals)

Thank you very much for helping!

[Valentina]

I have installed the --dev version and running the settings.xml (attached above). I get the following error:

csg_get_interaction_property: bondtype 'bonded' needs a mapping file (cg.inverse.map in xml) to determine the actual bond type (when calling from csg_call better use --ia-type bond, angle or dihedral)

But it is nit clear in the manual what format is it in:

• map Special mapping file for rdf calculations needed for bonded interactions

As I understand it is called within functions_common.sh
and in manual it says cg.inverse.map (optional)

I never had to use it before, could you advise me on it, please.

[Christoph Junghans]

2012/10/17 Valentina <valentina...@yahoo.co.uk>:

> I have installed the --dev version and running the settings.xml (attached
> above). I get the following error:
>
>
> csg_get_interaction_property: bondtype 'bonded' needs a mapping file
> (cg.inverse.map in xml) to determine the actual bond type (when calling from
> csg_call better use --ia-type bond, angle or dihedral)

You will have to specify the mapping file in cg.inverse.map.

The mapping file is just a 1:1 map with bond and angle definitions, see
http://code.google.com/p/votca/source/browse/hexane/ibi_bonded/hexane_cg.xml?repo=csg-tutorials
as an example.

The manual describes that somewhere in the section of csg_boltzmann.

After writing the mapping file I would run csg_stat once by hand with
the mapping file specified behind the --cg option.

Christoph

>
> But it is nit clear in the manual what format is it in:
>
> map Special mapping file for rdf calculations needed for bonded interactions
>
> As I understand it is called within functions_common.sh
> and in manual it says cg.inverse.map (optional)
>
> I never had to use it before, could you advise me on it, please.
>

> --
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/votca/-/tLVB8dNYKEQJ.

[Valentina]

Done, thank you. Now I am stuck with:

Using given table CC.pot.in for CC
################################################################################################################################
#                                                                                                                              #
# ERROR:                                                                                                                       #
# potential_extrapolate.sh: please specify add potential type (--type options) from: non-bonded bond thermforce angle dihedral #
# For details see the logfile /home/valentina/CG/inverse.log                                                                   #
#                                                                                                                              #
################################################################################################################################
Terminated


I am not sure where to add this. I have added like this to every interaction:

  <bonded>
    <name>CC</name>
    <min>0</min>
    <max>4</max>
    <step>0.01</step>

    <inverse>
      <target>CC.dist.tgt</target>
      <do_potential>1 0 0 0 </do_potential>

      <post_update>scale</post_update>
      <post_update_options>
       <scale>0.1</scale>
      </post_update_options>

      <post_update>pressure</post_update>
      <post_update_options>    
       <pressure>
        <type>simple</type>
        <do> 1 0 0 0</do>
        <simple>
          <scale>0.0001</scale>
        </simple>
       </pressure>
      </post_update_options>

      <post_update_options>
       <extrapolate> --type bond </extrapolate>
      </post_update_options>

      <post_add></post_add>
      <gromacs>
        <table>table_b1.xvg</table>
      </gromacs>
    </inverse>
  </bonded>

But it doesn't seem to pick up the input.

Also the programm asked me to provide CC.pot.in
I was not sure what is this potential, so I provided CC-bond.pot.ib (the output from csg_boltzmann) - is this right?

Thank you very much,
V

[Christoph Junghans]

2012/10/17 Valentina <valentina...@yahoo.co.uk>:

> Done, thank you. Now I am stuck with:
>
> Using given table CC.pot.in for CC
> ################################################################################################################################
> #
> #
> # ERROR:
> #
> # potential_extrapolate.sh: please specify add potential type (--type
> options) from: non-bonded bond thermforce angle dihedral #
> # For details see the logfile /home/valentina/CG/inverse.log
> #
> #
> #
> ################################################################################################################################
> Terminated

Something is not right. Have you tried to run the hexane ibi_bonded potential?
Just in case to check your installation.

Can you send me the last 100 lines of inverse.log?

Christoph

> --
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/votca/-/uahanZ2h6ccJ.

[Valentina]

calc thermforce calc_thermforce.sh
optimizer prepare_state optimizer_prepare_state.sh
optimizer parameters_to_potential optimizer_parameters_to_potential.sh
optimizer state_to_potentials optimizer_state_to_potentials.sh
optimizer state_to_mapping optimizer_state_to_mapping.sh
update optimizer update_optimizer.sh
update optimizer_single update_optimizer_single.sh
optimizer_target rdf optimizer_target_rdf.sh
optimizer_target density optimizer_target_density.sh
optimizer_target pressure optimizer_target_pressure.sh
simplex precede_state simplex_downhill_processor.pl
cma get cma_get.sh
cma precede_state cma_processor.py
calc target_rdf calc_target_rdf_generic.sh
pressure_cor simple pressure_cor_simple.pl
pressure_cor wjk pressure_cor_wjk.pl
kbibi ramp_correction kbibi_ramp_correction.pl
calc kbint calc_kbint.sh
density symmetrize density_symmetrize.py
table add add_POT.pl
table integrate table_integrate.pl
table extrapolate table_extrapolate.pl
table merge merge_tables.pl
table smooth table_smooth.pl
table linearop table_linearop.pl
table dummy table_dummy.sh
table get_value table_get_value.pl
table getsubset table_getsubset.py
table smooth_borders table_smooth_borders.py
table compare table_combine.pl --die --op =
table combine table_combine.pl
table average table_average.sh
table scale table_scale.pl
table change_flag table_change_flag.sh
table functional table_functional.sh
potential extrapolate potential_extrapolate.sh
configuration compare configuration_compare.py
tables jackknife tables_jackknife.pl
run gromacs run_gromacs.sh
presimulation gromacs run_gromacs.sh --pre
pressure gromacs calc_pressure_gromacs.sh
rdf gromacs calc_rdf_generic.sh
imc_stat gromacs imc_stat_generic.sh
density gromacs calc_density_gromacs.sh
prepare_generic gromacs prepare_generic_gromacs.sh
initstep_generic gromacs initialize_step_generic_gromacs.sh
prepare_generic espresso prepare_generic_espresso.sh
initstep_generic espresso initialize_step_generic_espresso.sh
convert_potential gromacs potential_to_gromacs.sh
convert_potential xvg table_to_xvg.pl
functions gromacs functions_gromacs.sh
run espresso run_espresso.sh
pressure espresso calc_pressure_espresso.sh
rdf espresso calc_rdf_espresso.sh
convert_potential espresso potential_to_espresso.sh
convert_potential tab table_to_tab.pl
functions espresso functions_espresso.sh
------------------------
Prepare (dir step_000)
------------------------
cp_from_main_dir: 'grompp.mdp topol.top table_C_C.xvg table_a1.xvg table_b1.xvg table_d1.xvg  index.ndx'
Running critical command 'pushd /home/valentina/CG'
~/CG ~/CG/step_000
Running critical command 'cp grompp.mdp topol.top table_C_C.xvg table_a1.xvg table_b1.xvg table_d1.xvg index.ndx /home/valentina/CG/step_000'
Running critical command 'popd'
~/CG/step_000
Running subscript 'prepare_generic.sh' (from tags prepare ibi) dir /share/apps/votca/share/votca/scripts/inverse
For all bonded
for_all: run 'do_external prepare_single ibi'
Running subscript 'prepare_generic_single.sh' (from tags prepare_single ibi) dir /share/apps/votca/share/votca/scripts/inverse
Running critical command 'sed -n s/.*cg_bonded\.\([^[:space:]]*\) .*/\1/p'

Using given table CC.pot.in for CC

Running critical command 'mktemp CC.pot.in.smooth.XXX'
Converting /home/valentina/CG/CC.pot.in to CC.pot.new
Running critical command 'csg_resample --in /home/valentina/CG/CC.pot.in --out CC.pot.in.smooth.sU6 --grid 0:0.01:4 --comment Created on Wed Oct 17 17:42:27 BST 2012 by vale...@compute-0-4.local
called from prepare_generic_single.sh, version 1.3-dev hgid: cae513aba385
settings file: /home/valentina/CG/settings.xml
working directory: /home/valentina/CG/step_000'
Running critical command 'mktemp CC.pot.in.extrapolate.XXX'
Running subscript 'potential_extrapolate.sh --type CC.pot.in.smooth.sU6 CC.pot.in.extrapolate.YdY' (from tags potential extrapolate) dir /share/apps/votca/share/votca/scripts/inverse

Callstack:
/share/apps/votca/share/votca/scripts/inverse/inverse.sh - linenumber 166
    do_external - linenumber 170 in /share/apps/votca/share/votca/scripts/inverse/functions_common.sh
        /share/apps/votca/share/votca/scripts/inverse/prepare_generic.sh - linenumber 31
            for_all - linenumber 17 (see 'csg_call --cat function for_all')
                    do_external - linenumber 18 (see 'csg_call --cat function do_external')
                        /share/apps/votca/share/votca/scripts/inverse/prepare_generic_single.sh - linenumber 43
                            do_external - linenumber 18 (see 'csg_call --cat function do_external')
                                /share/apps/votca/share/votca/scripts/inverse/potential_extrapolate.sh - linenumber 64
                                    die - linenumber 2 (see 'csg_call --cat function die')

################################################################################################################################
#                                                                                                                              #
# ERROR:                                                                                                                       #
# potential_extrapolate.sh: please specify add potential type (--type options) from: non-bonded bond thermforce angle dihedral #
# For details see the logfile /home/valentina/CG/inverse.log                                                                   #
#                                                                                                                              #
################################################################################################################################

die: (called from 591)  CSG_MASTER_PID is 32479
die: pids to kill: 0

On Friday, September 28, 2012 2:07:44 PM UTC+1, Valentina wrote:

[Valentina]

No, I have not run hexane ibi_bonded. I have not realised there is a tutorial for it. I may looking in wrong place - but I dont seem to find it in votca-csg-tutorials-1.2.3.tar.gz 
am I looking in the wrong place? Or shall I write ibi_bonded myself?

Thanks
V

[Christoph Junghans]

2012/10/17 Valentina <valentina...@yahoo.co.uk>:

> No, I have not run hexane ibi_bonded. I have not realised there is a
> tutorial for it. I may looking in wrong place - but I dont seem to find it
> in votca-csg-tutorials-1.2.3.tar.gz
> am I looking in the wrong place? Or shall I write ibi_bonded myself?

You can get it via the buildutil

./build --dev csg-tutorials

Christoph

> https://groups.google.com/d/msg/votca/-/MN8M51uySo0J.

[Valentina]

Ok, great. I will try the tutorial and see where I have gone wrong.

[Christoph Junghans]

Hi Valentina,

I had another look at your input files and found the problem.

- The name you give a bonded interaction in the mapping file has to be
the same as the name of the interaction in the settings file. That is
how the type detection works. (There was a check missing in the code,
which lead to error in the extrapolation, I fixed that. You can use
"./build -du csg" to get the update)

- In your mapping file you have defined the bonds and angle, but the
definition of the beads is missing. Have a look the hexane example, in
principle you have to create a 1:1 map. And try to run csg_stat with
--cg option once by hand to see if the distributions are calculated
correctly.

Christoph

2012/10/17 Valentina <valentina...@yahoo.co.uk>:

> https://groups.google.com/d/msg/votca/-/JWRcDYtHxR8J.

[Valentina Erastova]

On 18 Oct 2012, at 11:49, Christoph Junghans wrote:

> Hi Valentina,
>
> I had another look at your input files and found the problem.
>
> - The name you give a bonded interaction in the mapping file has to be
> the same as the name of the interaction in the settings file. That is
> how the type detection works. (There was a check missing in the code,
> which lead to error in the extrapolation, I fixed that. You can use
> "./build -du csg" to get the update)

Great, updating now.

>
> - In your mapping file you have defined the bonds and angle, but the
> definition of the beads is missing. Have a look the hexane example, in
> principle you have to create a 1:1 map. And try to run csg_stat with
> --cg option once by hand to see if the distributions are calculated
> correctly.

WIll do that.

Thank you for your help!

[Valentina]

Hello,

I am having a problem, I cannot identify how to solve.

I am doing IBI of bonded interactions and so I get an error.

There must be smth wrong between cg-mapping.xml and topol.top files, but I cannot spot what. I also get  error of same type if I run

 csg_dump as a check.

Thank you!



List of molecules:
--------------------------------------
WARNING: unknown molecule "DCG" with id 0 in topology
Reading frame       0 time    0.000   molecule will not be mapped to CG representation
Check weather a mapping file for all molecule exists, was specified in --cg separated by ; and the ident tag in xml-file matches the molecule name
--------------------------------------

Reading frame      50 time   10.000   an error occurred:
Topology does not have beads of type "C"
This was specified in type1 of interaction "C-C"

Callstack:
/share/apps/votca/share/votca/scripts/inverse/inverse.sh - linenumber 252

    do_external - linenumber 170 in /share/apps/votca/share/votca/scripts/inverse/functions_common.sh

        /share/apps/votca/share/votca/scripts/inverse/update_ibi.sh - linenumber 31

            for_all - linenumber 17 (see 'csg_call --cat function for_all')
                    do_external - linenumber 18 (see 'csg_call --cat function do_external')

                        /share/apps/votca/share/votca/scripts/inverse/calc_rdf_generic.sh - linenumber 71
                            critical - linenumber 4 (see 'csg_call --cat function critical')

                                die - linenumber 2 (see 'csg_call --cat function die')

#########################################################################################################################################################################################

#                                                                                                                                                                                       #
# ERROR:                                                                                                                                                                                #

# critical: 'csg_stat --nt 8 --options /home/valentina/CG/GAS/settings.xml --top topol.tpr --trj traj.xtc --begin 10 --first-frame 0 --cg /home/valentina/CG/GAS/cg-mapping.xml' failed #
# For details see the logfile /home/valentina/CG/GAS/inverse.log                                                                                                                        #
#                                                                                                                                                                                       #
#########################################################################################################################################################################################
die: (called from 7667)  CSG_MASTER_PID is 7125

die: pids to kill: 0

On Friday, September 28, 2012 2:07:44 PM UTC+1, Valentina wrote:

[Christoph Junghans]

Hi,

2012/10/24 Valentina <valentina...@yahoo.co.uk>:

>
> Hello,
>
> I am having a problem, I cannot identify how to solve.
>
> I am doing IBI of bonded interactions and so I get an error.
>
> There must be smth wrong between cg-mapping.xml and topol.top files, but I
> cannot spot what. I also get error of same type if I run
>
> csg_dump as a check.

Could the identifier in the mapping xml file be wrong?
The identifier is the name from the [ moleculetype ] block in topol.top

If you have multiple molecule type in your system, you will need a
mapping file for each them, but I haven't added multiple type support
to csg_inverse.
It is not hard to add.

Christoph

> --
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/votca/-/6pOJ6qExAD4J.