different distribution when using cg mapping file.

[Chandan Choudhury]

Dear votca users,

I am iterating the non-bonded part of the polymer (with following sequence:

A1-B1-C1-B2-D1-B3-C2-B4-D2-B5-C3-E1). In step_001, my observations are following:

csg_stat uses the following command to compute the distributions :

csg_stat --nt 8 --options /data/home/chandan/Documents/votca/polycarbonate/trimer/mapping_II/re/ibi/nb/B_D/settings.xml --top topol.tpr --trj traj.xtc --begin 20 --first-frame 0

I obtain the distributions in B-D.dist.new file. Here in settings.xml, the min is 0.37

When, I change the min to 0. It gives the exact distribution.

But, when I try to include the information of all the other bonded and non-bonded interaction (min=0.0 and max values), then I need to use a mapping file (attached).
csg_stat --nt 8 --options settings_csg_stat.xml --top ../topol.tpr --trj ../traj.xtc --begin 20 --first-frame 0 --cg ../../pc_new_cg.xml

Here, the setting contains all bonded and non-bonded part with min=0.0. The distribtution I obtain is B-D.dist_withpc-xml.new.

I also tried the g_rdf tool (with nrexcl=3). The output file is B-D.rdf.xvg. Here the selection are between all B's and D's.

It can be seen that the distribution obtained are different. On using coarse-grained xml and when I donot use it, the distributions are different. The gmx tool also gives the different result.

Am I doing something wrong. I can suspect the problem might be in the cg mapping file. But, I am unable to locate it. Can some one help me.

The trajectory and tpr is same for all the above commands.

Chandan

--
Chandan kumar Choudhury
NCL, Pune
INDIA

[Christoph Junghans]

2013/9/30 Chandan Choudhury <iit...@gmail.com>:

The trajectory is missing, just send a conf.gro. One frame will be
enough to check for the error.

These are two separate issues.

1.) The difference between g_rdf and csg_stat without mapping file
Could you check that all exclusion you expect, show up in:
$ csg_stat --top topol.tpr --execl
and compare to the exclusion in
$ gmxdump -s topol.tpr

2.) The difference between csg_stat with and without mapping file
For this you have to understand how exclusion in VOTCA.
If you don't specify a mapping file, exclusion are just read for the
topology (tpr).
If you specify a mapping file, exclusion are generated by VOTCA due to
the following rule.
Everything which is connected via a bond, angle or dihedral is excluded.
In your case to generate exclusion equivalent to nrexcl=3, you will
need add some extra dihedrals to the mapping file (with evaluating
them).

If you are unsure compare:
$ csg_stat --top topol.tpr --execl
with
$ csg_stat --top topol.tpr --execl --cg pc_new_cg.xml

We need to document that better, patch for the manual are welcome!
<https://code.google.com/p/votca/wiki/Manual>

Cheers,

Christoph

>
> Chandan
> --
> Chandan kumar Choudhury
> NCL, Pune
> INDIA
>

> --
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to votca+un...@googlegroups.com.
> To post to this group, send email to vo...@googlegroups.com.
> Visit this group at http://groups.google.com/group/votca.
> For more options, visit https://groups.google.com/groups/opt_out.



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

[Chandan Choudhury]

On Tue, Oct 1, 2013 at 2:13 AM, Christoph Junghans <jung...@votca.org> wrote:

The  trajectory is missing, just send a conf.gro. One frame will be
enough to check for the error.

 

I have attached conf.gro 

These are two separate issues.

1.) The difference between g_rdf and csg_stat without mapping file
Could you check that all exclusion you expect, show up in:
$ csg_stat --top topol.tpr --execl
and compare to the exclusion in
$ gmxdump -s topol.tpr 

csg_dump --excl --top ../topol.tpr > excl_csg 

csg_dump --excl --top ../topol.tpr --cg pc_new_cg.xml > excl_csg_pc

The exclusion list for csg_dump with and without pc_new_cg.xml are same. I could not clearly understand the exclusions of the gromacs.

I have attached the output of the csg_dump and gmxdump. 

2.) The difference between csg_stat with and without mapping file
For this you have to understand how exclusion in VOTCA.
If you don't specify a mapping file, exclusion are just read for the
topology (tpr).
If you specify a mapping file, exclusion are generated by VOTCA due to
the following rule.
Everything which is connected via a bond, angle or dihedral is excluded.
In your case to generate exclusion equivalent to nrexcl=3, you will
need add some extra dihedrals to the mapping file (with evaluating
them).

If you are unsure compare:
$ csg_stat --top topol.tpr --execl
with
$ csg_stat --top topol.tpr --execl --cg pc_new_cg.xml

I could understand that. Thanks for pointing out. On supplying the pc_new_cg.xml to csg_dump the excl matches.

When I execute :
csg_stat --nt 8 --options settings.xml --top ../topol.tpr --trj ../traj.xtc --begin 20 --first-frame 0 --cg pc_new_cg.xml

I obtain only the B-D.dist.new (as the settings.xml contains the information about the B-D.)

When I perform :

csg_stat --nt 8 --options settings_all.xml --top ../topol.tpr --trj ../traj.xtc --begin 20 --first-frame 0 --cg pc_new_cg.xml

I obtain all the bonded and non-bonded distributions. B-D.dist_all.new. 

In both the cases the pc_new_cg has been updated with the inclusion of dihedral information. 

It can be seen that the distributions differ.

Only difference between the above two steps is the information in the settings file. 

What I understand to obtain the correct distributions we should use all the non-bonded information in the settings file. Please correct me if I am wrong.

We need to document that better, patch for the manual are welcome!
<https://code.google.com/p/votca/wiki/Manual>

Are you suggesting me to incorporate patches to the manual. If so, I followed the above link but could not figure out how can I do it. 

Chandan

[Christoph Junghans]

2013/10/1 Chandan Choudhury <iit...@gmail.com>:

There is a typo in your settings.xml, type2=C instead of D.
Anyhow after changing that I did:
$ csg_stat --nt 8 --top topol.tpr --trj conf.gro --options
settings.xml --cg pc_new_cg.xml
$ csg_stat --nt 8 --top topol.tpr --trj conf.gro --options
settings_all.xml --cg pc_new_cg.xml --ext dist_all.new
and
$ echo -e "a B?\na D?\nq" | make_ndx -f conf.gro
$ echo -e "B?\nD?" | g_rdf -s topol.tpr -f conf.gro -n index.ndx -bin 0.01
If you now look at plot
gnuplot> p "rdf.xvg" w l,"B-D.dist.new" w l, "B-D.dist_all.new" w l

you will see that rdf.xvg, B-D.dist.new and B-D.dist_all.new exactly the same.

I am not sure where your original difference came from. The only
things, which might make a difference is the "--begin 20", but you
can do the same with the -"b- option of g_rdf.

>
>>
>> We need to document that better, patch for the manual are welcome!
>> <https://code.google.com/p/votca/wiki/Manual>
>>
> Are you suggesting me to incorporate patches to the manual. If so, I
> followed the above link but could not figure out how can I do it.

The "Contributing a change" paragraph tells you how to submit updates
to the manual.
As you have already built the manual in past, it should be an easy thing.

Cheers,

Christoph
>
> Chandan

[Chandan Choudhury]

On Tue, Oct 1, 2013 at 8:34 PM, Christoph Junghans <jung...@votca.org> wrote:

There is a typo in your settings.xml, type2=C instead of D.
Anyhow after changing that I did:

Thanks. It worked for me now.

$ csg_stat --nt 8 --top  topol.tpr --trj conf.gro --options
settings.xml --cg pc_new_cg.xml
$ csg_stat --nt 8 --top  topol.tpr --trj conf.gro --options
settings_all.xml --cg pc_new_cg.xml --ext dist_all.new
and
$ echo -e "a B?\na D?\nq" | make_ndx -f conf.gro
$ echo -e "B?\nD?" | g_rdf -s topol.tpr -f conf.gro -n index.ndx -bin 0.01
If you now look at plot
gnuplot> p "rdf.xvg" w l,"B-D.dist.new" w l, "B-D.dist_all.new" w l

you will see that rdf.xvg, B-D.dist.new and B-D.dist_all.new exactly the same.

I am not sure where your original difference came from. The only
things, which might make a difference is the  "--begin 20", but you
can do the same with the -"b- option of g_rdf.

The distribution matches very well.

>
>>
>> We need to document that better, patch for the manual are welcome!
>> <https://code.google.com/p/votca/wiki/Manual>
>>
> Are you suggesting me to incorporate patches to the manual. If so, I
> followed the above link but could not figure out how can I do it.

The "Contributing a change" paragraph tells you how to submit updates
to the manual.
As you have already built the manual in past, it should be an easy thing.

I will give a try.

Cheers,

Christoph

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

--
You received this message because you are subscribed to the Google Groups "votca" group.
To unsubscribe from this group and stop receiving emails from it, send an email to votca+un...@googlegroups.com.
To post to this group, send email to vo...@googlegroups.com.
Visit this group at http://groups.google.com/group/votca.
For more options, visit https://groups.google.com/groups/opt_out.

Chandan

[Christoph Junghans]

2013/10/1 Chandan Choudhury <iit...@gmail.com>:

>
>
> On Tue, Oct 1, 2013 at 8:34 PM, Christoph Junghans <jung...@votca.org>
> wrote:
>>
>> There is a typo in your settings.xml, type2=C instead of D.
>> Anyhow after changing that I did:
>
> Thanks. It worked for me now.
>>
>> $ csg_stat --nt 8 --top topol.tpr --trj conf.gro --options
>> settings.xml --cg pc_new_cg.xml
>> $ csg_stat --nt 8 --top topol.tpr --trj conf.gro --options
>> settings_all.xml --cg pc_new_cg.xml --ext dist_all.new
>> and
>> $ echo -e "a B?\na D?\nq" | make_ndx -f conf.gro
>> $ echo -e "B?\nD?" | g_rdf -s topol.tpr -f conf.gro -n index.ndx -bin 0.01
>> If you now look at plot
>> gnuplot> p "rdf.xvg" w l,"B-D.dist.new" w l, "B-D.dist_all.new" w l
>>
>> you will see that rdf.xvg, B-D.dist.new and B-D.dist_all.new exactly the
>> same.
>>
>> I am not sure where your original difference came from. The only
>> things, which might make a difference is the "--begin 20", but you
>> can do the same with the -"b- option of g_rdf.
>
> The distribution matches very well.

Also if I change min to 0.45, the distribution still match exactly.
So I cannot really reproduce what happens for you before.

[Chandan Choudhury]

On Wed, Oct 2, 2013 at 3:22 AM, Christoph Junghans <jung...@votca.org> wrote:

2013/10/1 Chandan Choudhury <iit...@gmail.com>:

>
>
> On Tue, Oct 1, 2013 at 8:34 PM, Christoph Junghans <jung...@votca.org>
> wrote:
>>
>> There is a typo in your settings.xml, type2=C instead of D.
>> Anyhow after changing that I did:
>
> Thanks. It worked for me now.
>>
>> $ csg_stat --nt 8 --top  topol.tpr --trj conf.gro --options
>> settings.xml --cg pc_new_cg.xml
>> $ csg_stat --nt 8 --top  topol.tpr --trj conf.gro --options
>> settings_all.xml --cg pc_new_cg.xml --ext dist_all.new
>> and
>> $ echo -e "a B?\na D?\nq" | make_ndx -f conf.gro
>> $ echo -e "B?\nD?" | g_rdf -s topol.tpr -f conf.gro -n index.ndx -bin 0.01
>> If you now look at plot
>> gnuplot> p "rdf.xvg" w l,"B-D.dist.new" w l, "B-D.dist_all.new" w l
>>
>> you will see that rdf.xvg, B-D.dist.new and B-D.dist_all.new exactly the
>> same.
>>
>> I am not sure where your original difference came from. The only
>> things, which might make a difference is the  "--begin 20", but you
>> can do the same with the -"b- option of g_rdf.
>
> The distribution matches very well.

Also if I change min to 0.45, the distribution still match exactly.
So I cannot really reproduce what happens for you before.

Thanks for troubleshooting. 

Few things were unknowingly mistaken.

1. I did not use the information of dihedrals in the exclusions. The exclusion list was different. This might have added to the different distribution.

2. The typo you figured out. Thanks for this. I now crossed checked with changing the min, it did reproduce the distribution.

Thanks a lot once again.

Chandan