[gmx-users] g_spatial Problem

Wed Oct 5 05:45:36 CEST 2016

Dear Dr. Neale,

Thanks for your reply. Actually my system is quite a big system. It
contains:

Cation - 250 molecules
Anion - 250 molecules
Type A  (urea)- 500 molecules
Type B (refrigerant)- 250 molecules

I want to get spatial distribution function like the following. I used this
picture as because this system is similar to mine:
https://s26.postimg.org/hmw56k5d5/demo.jpg

Let's say I am trying to get the sdf of anions around the cation (solute).
When I am going through the manual of g_spatial, it did not specify I have
to choose a single atom of solute. So first I took all the molecules of
cation as a centered group and followed the steps that I mentioned in the
earlier mail. I got something like this:
https://s26.postimg.org/4794h3wvd/image.jpg
If I load both grid files, it turned out like this :
https://s26.postimg.org/hpg0te90p/image.jpg

First of all it's rectangular, and all the sdf that I saw in the paper are
spherical and in the center there are only one molecule (which is clearly
different from mine).

In the forum, I read that there should be only one centered molecule. So I
made an index file including a random molecule of Cation and followed the
steps. This time I got the image like this:
https://s26.postimg.org/7gs51bacp/25percent_Copy.jpg

So I am not sure what I am doing wrong. I tried to look for help. But I
could not find any till now. Right now I am completely stuck.

The process you said- to make 10 separate trajectories, in my case there
are 250 molecules of solute and I have to do 250 analysis and then
concatenate all the trajectories? I am not clear about this. I have
completed my simulation, I have the  final trjectory and gro file. What you
are suggesting is to take the trajectory and output it as 1 solute + all
solvents (anion + type A + type B) and then concatenate all these
trajectories.  Then run the steps that I enlisted. It is something like
this?

I am really thankful for your time.

Rubaiyet

On Tue, Oct 4, 2016 at 9:44 PM, Christopher Neale <
chris.neale at alum.utoronto.ca> wrote:

> Can you please describe the system a bit more (especially how many
> molecules of each of the 4 types you have) and also provide more
> information about what you want to obtain (sdf, obviously, but of what and
> ideally also why).
>
> 1. your procedure could be ok, but also maybe not. Depends on what you are
> trying to achieve. I am presuming there is only 1 "anion" though (based on
> the image you sent me off-list) although that doesn't jive with your
> question #2, so I am confused here. Fitting over multiple "anion" molecules
> that are separately diffusing is not likely what you want.... -- (B) the
> grid.cube file is a rectangular matrix, so that is expected. The regions
> with non-zero density will not fill up the entire rectangle though, usually.
>
> 2. The g_spatial program is intended for a single central fitting group.
> There was an old program called g_sdf that may (or may not) do what you
> want -- it was intended to look at multiple central solutes though as I
> recall. If you want to have the cumulative sdf over all "anions" then you
> have to do more processing. What I do in this case is this. Say I have 10
> "anion" solute molecules (the central group for which you want to build an
> sdf of other molecules around ... I'm going to call this the solute from
> here on). Then you make 10 copies of the trajectory in which you only
> output one solute (different one each time) and also include all the
> solvent (this part will be the same in all trajectories). Then concatenate
> all of these trajectories with trjconv -cat -nosort. You may need to create
> a .tpr for this trajectory, or possibly g_spatial will accept a .gro or
> .pdb as input to the -s option. Now you take this concatenated trajectory
> and run it through all the steps that you listed in your email. This should
> give you what you want.
>
> Note: if your solute is flexible, then you may not want to fit on the
> entire solute. That may lead to smearing. I would certainly try it, but I
> would also try fitting on different subsets of the solute (like one charged
> region) that have at least 3 atoms and make that sdf as well.
>
> Good luck.
> Chris.
>
> ________________________________________
> From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se <
> gromacs.org_gmx-users-bounces at maillist.sys.kth.se> on behalf of Rubaiyet
> Abedin <abedin.r at husky.neu.edu>
> Sent: 26 September 2016 13:18:46
> To: gromacs.org_gmx-users at maillist.sys.kth.se
> Subject: [gmx-users] g_spatial Problem
>
> Dear Sir,
>
> Hope you are doing good. I am working with a system that contains ionic
> liquid and refrigerant. To avoid complication let’s say my system contains
> the following:
>
> Cation (1)
>
> Anion (2)
>
> Type A (3)
>
> Type B (4)
>
>
>
> And the system (0). I want to get the spatial distribution function of type
> A around Anion. So the central molecule will be the anion. I used the
> following commands:
>
>
>
> *1.      **make_ndx -f bath_298K.part0001.gro -o bath_298K_index.ndx*
>
>
>
> *2.      **trjconv -s bath_298K.tpr -f bath_298K.part0001.xtc -o
> bath_298K_b.xtc -center  -ur compact -pbc none  -n bath_298K_index.ndx*
>
>
>
> I selected Anion (2) for centering and system (0) for output
>
>
>
> *3.      **trjconv -s bath_298K.tpr -f bath_298K_b.xtc -o bath_298K_c.xtc
> -fit rot+trans*
>
>
>
> I selected Anion (2) for fitting and system (0) for output
>
>
>
> *4.      **g_spatial -s bath_298K.tpr -f bath_298K_c.xtc -n
> bath_298K_index.ndx -nab 100*
>
>
>
> I selected Anion (2) to generate SDF and type A (3) to output coordinates
>
>
>
> *5.      **g_spatial -s bath_298K.tpr -f bath_298K_c.xtc -n
> bath_298K_index.ndx -nab 100*
>
>
>
> I selected Anion (2) to generate SDF and Anion (2) to output coordinates
>
>
>
> I got two grid.cube file and load them into vmd. For the grid.cube (from
> step 4) I chose the isosurface value 0.01 and viewed the other cube file as
> a single molecule. I have several question:
>
>
>
> 1.      Do you think I am following the right procedure specially I am
> confused and step 4 and 5. After loading both of the grid.cube file I got
> something like this the attached.
>
>  For some cases I even got rectangular box. Should it be a sphere?
>
> 2.      I am choosing a random Anion (2). Every time I choose a different
> anion, I see different cloud shape. I am confused if I am doing something
> wrong or I have to choose the central molecule following any specific
> technique.
>
>
> Please help me out with this. Thanks a lot.
>
>
>
> Rubaiyet
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/
> Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>