How to exclude bonded atoms from the distribution calculations?

[Sergey Larin]

Dear votca users,

I've faced one more problem (not so big, I suppose) in coarsegrainig using votca. I've found that csg_stat calculate pair distribution functions taking into account atoms that are bonded. As a result the distributions have peaks that correspond to bonded interaction (see the picture, in the molecule I'm coarsegraining the monomer unit is -C-D-A-D-C-A-B-B-A-B-). It is clear for me that it is pretty useless to perform IBI for non-bonded interactions using these distribution as target ones. Thus, my question is how to say csg_stat to exclude bonded atoms from distribution caclulations? Unfortunately, I've found only the way to create exclusion list for non-bonded interactions to perform reference simulations for bonded interactions distributions.

Thanks in advance.

Sincerely,
Sergey

[Sikandar Mashayak]

Hi Sergey

From your description of the problem, I understand that you want to use csg_stat to compute
only the non-bonded distributions between specified atoms. Right?

If so.. csg_stat can be used to compute both bonded and non-bonded distributions separately.

To achieve that one approach would be to specify your molecule topology that each atom has

unique and different name though there may be multiple atoms of same type. In your monomer case,

there are four different types of atoms A, B, C, D, and each monomer consists of 2 As, 3 Bs, 2Cs, and 2 Ds.
Then, in topology, you can specify atom types as A, B, C, D and atom names are A1, A2, B1, B2, B3, C1, C2,
and D1, and D2. For clarification, you can check out the propane example in votca-tutorials.
 

Now with above mentioned topolopy, you can compute non-bonded distributions between A1-A2, C1-C2, and so on.

Next, if you want to perform coarse-graining and need a reference pair distributions then first you need to define
a mapping to specify how to determine CG beads configurations from a given reference all-atom configuration. Then

you can provide this mapping file to csg_stat to determine target distributions for coarse-graining.

Again you can check out propane example and votca manual to get more information about how to achieve this.

I hope this helps.

Cheers
Sikandar

--
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.

[Sergey Larin]

Dear Sikandar,

Thank you for your help. I've done everything you describe already. I've prepared mapping and topology files and start test IBI process. I'll try to describe problem that exists, from my point of view in more details.
Let's use notation proposed by you and see at A-B distribution. Monomer unit of my polymer consists of 11 beads that have 4 different types: -C1-D1-A1-D1-C1-B1-A2-B2-B3-A3-B4-
This topology is described in VOTCA mapping file that I use to calculate distributions with csg_stat. Of course, it excludes interactions for beads bonded directly. for example, A2-B2 and B3-A3 pairs are not included in distribution. This is right. But it should also exclude pairs in which beads are connected by two consecutive bonds, for example, A2-B3 and B2-A3, but csg_stat does not do this exclusion. That is why I have the peak on the A-B pair distribution function. And this is a problem.
As far as I know, Gromacs does not calcuate non-bonded interactions for beads connected by one, two and three consecutive bonds, thus only interactions between pairs B1-A3 and A2-B4 are calculated during simulation. Thus, only these two pairs should be taken into account when calculation distribution, and csg_stat does not do this.

So, my question is how to exclude 1-3 and 1-4 pairs from distributions calculated by csg_stat? Is it possible at all? Unfortunately I've not found any solutions neither in manual, nor in tutorial samples.

Sincerely,
Sergey


суббота, 14 декабря 2013 г., 0:41:51 UTC+4 пользователь sikandar написал:

[Christoph Junghans]

2013/12/16 Sergey Larin <svla...@gmail.com>

>
> Dear Sikandar,
>
> Thank you for your help. I've done everything you describe already. I've prepared mapping and topology files and start test IBI process. I'll try to describe problem that exists, from my point of view in more details.
> Let's use notation proposed by you and see at A-B distribution. Monomer unit of my polymer consists of 11 beads that have 4 different types: -C1-D1-A1-D1-C1-B1-A2-B2-B3-A3-B4-
> This topology is described in VOTCA mapping file that I use to calculate distributions with csg_stat. Of course, it excludes interactions for beads bonded directly. for example, A2-B2 and B3-A3 pairs are not included in distribution. This is right. But it should also exclude pairs in which beads are connected by two consecutive bonds, for example, A2-B3 and B2-A3, but csg_stat does not do this exclusion. That is why I have the peak on the A-B pair distribution function. And this is a problem.
> As far as I know, Gromacs does not calcuate non-bonded interactions for beads connected by one, two and three consecutive bonds, thus only interactions between pairs B1-A3 and A2-B4 are calculated during simulation. Thus, only these two pairs should be taken into account when calculation distribution, and csg_stat does not do this.
>
> So, my question is how to exclude 1-3 and 1-4 pairs from distributions calculated by csg_stat? Is it possible at all? Unfortunately I've not found any solutions neither in manual, nor in tutorial samples.

VOTCA has no way to define the number of consecutive bonds to exclude,
but you can create a couple of extra bonds and angles to achieve the
same number of exclusions.

Cheers,

Christoph

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

[Sergey Larin]

понедельник, 16 декабря 2013 г., 20:04:38 UTC+4 пользователь Christoph Junghans написал:

2013/12/16 Sergey Larin <svla...@gmail.com>
>
> Dear Sikandar,
>
> Thank you for your help. I've done everything you describe already. I've prepared mapping and topology files and start test IBI process. I'll try to describe problem that exists, from my point of view in more details.
> Let's use notation proposed by you and see at A-B distribution. Monomer unit of my polymer consists of 11 beads that have 4 different types: -C1-D1-A1-D1-C1-B1-A2-B2-B3-A3-B4-
> This topology is described in VOTCA mapping file that I use to calculate distributions with csg_stat. Of course, it excludes interactions for beads bonded directly. for example, A2-B2 and B3-A3 pairs are not included in distribution. This is right. But it should also exclude pairs in which beads are connected by two consecutive bonds, for example, A2-B3 and B2-A3, but csg_stat does not do this exclusion. That is why I have the peak on the A-B pair distribution function. And this is a problem.
> As far as I know, Gromacs does not calcuate non-bonded interactions for beads connected by one, two and three consecutive bonds, thus only interactions between pairs B1-A3 and A2-B4 are calculated during simulation. Thus, only these two pairs should be taken into account when calculation distribution, and csg_stat does not do this.
>
> So, my question is how to exclude 1-3 and 1-4 pairs from distributions calculated by csg_stat? Is it possible at all? Unfortunately I've not found any solutions neither in manual, nor in tutorial samples.
VOTCA has no way to define the number of consecutive bonds to exclude,
but you can create a couple of extra bonds and angles to achieve the
same number of exclusions.

O'K. Thanks. Could you say which atoms are excluded by csg_stat? Connencted by bonds and included in angles and dihedrals? If it so, I'll define additional dihedral angles in my simulation and it should solve the problem.

Sergey

[Christoph Junghans]

2013/12/16 Sergey Larin <svla...@gmail.com>:

We follow the Gromacs convention, bonds and angles and dihedrals will
automatically generate exclusion, so adding angles will be fine.
If you use csg_stat without a mapping file the exclusions will be read
from the tpr file, but a mapping reset all exclusions.

To check your exclusions, you can do something like:
$ csg_dump --top topol.tpr --cg mapping.xml --excl

Cheers,

Christoph

[Sergey Larin]

Thank you, Cristoph.

I've added dihedrals to mapping and now it seems to be OK.

Best wishes,
Sergey

вторник, 17 декабря 2013 г., 20:26:27 UTC+4 пользователь Christoph Junghans написал: