[gmx-users] rdf

[chris.neale at utoronto.ca]

make a figure of your rdf, post it online somewhere (I use  
photobucket) and reply to the list with a link to your figure. Make  
sure your figure is well labeled and give us a thorough description of  
what you see that you don't like. Also tell us what your unit-cell box  
vectors are. Also describe your system thoroughly, e.g. one methane in  
a box of 100 water molecules...

I know that the meaning of "values fluctuate alot (i.e. decreasing)."  
is immediately obvious to you, but for those of us who haven't seen  
the graph it is pretty tough to figure out what is actually going on.

And off the bat, I'd say that the tail of your RDF is just poorly  
defined (data coming from few waters in the figurative corners of the  
unit cell for a central methane) and you are seeing the noise that  
results from this -- Try regenerating your data by block averaging  
into 0-33 ns, 33-66 ns, and 66-100 ns, then plot the average and  
standard deviation of your results... I bet this will show you what's  
going on.

Chris.

> Hi Will,
>
>     I have been using pme from the very start and it didn't help  
> with the volume. Could you please elaborate on what you mean by  
> average g(r) over every molecule?
>
>    And because my rdf-s are not smooth enough even with 100ns, I get  
> fluctuating data using rdf-s.
>

I think this is a case where it is important to refer back to previous posts.
Is this related to your system that has one molecule of methane solvated in
water?  If I recall, Chris already told you why the values you see  
aren't smooth:

http://lists.gromacs.org/pipermail/gmx-users/2010-January/048297.html

Related to the volume issue, what type of pressure coupling are you applying?
Is your treatment of van der Waals appropriate (have you applied the switching
function properly)?  Is switch appropriate for your chosen force  
field?  Are the
cutoff's you are using those prescribed by this force field?  What is the
magnitude of the decrease in the volume?  Is it systematic?

It is also important to be clear when posting - you are using "vdwtype  
= switch"
and not "coulombtype = switch", hence the potential confusion when you say you
are just using switch.  It makes a big difference.  It is often most  
convenient
to post your .mdp file instead of leaving it to guesswork.

-Justin

> -Nisha
>
>
>
>
> Quoting Will Glover <will_glover at yahoo.com>:
>
>> Hi,
>> 100 ns is overkill for those liquids.  It should be smooth enough  
>> at  1 ns, provided you average g(r) over every molecule.
>> What do you mean the volume decreases?  With time?
>>
>> As for the cut-off, it's well known that switching Coulomb terms   
>> leads to artifacts, and increasing the cut-off distance doesn't   
>> necessarily make things converge.  See   
>> http://pubs.acs.org/doi/abs/10.1021/ct0502256 for example.
>>
>> Use PME for electrostatics.
>>
>> Regards,
>> -- 
>> Will
>>
>> --- On Sat, 1/30/10, nishap.patel at utoronto.ca  <nishap.patel at  
>> utoronto.ca> wrote:
>>
>>> From: nishap.patel at utoronto.ca <nishap.patel at utoronto.ca>
>>> Subject: [gmx-users] rdf
>>> To: "Discussion list for GROMACS users" <gmx-users at gromacs.org>
>>> Date: Saturday, January 30, 2010, 9:45 PM
>>> Hi,
>>>
>>>   I am doing rdf's of simple molecules. I ran my
>>> simulation of water and methane for 100ns to get a smoother
>>> curve for the rdf. I am trying to determine the volume, and
>>> after a certain cut-off I would assume my values to be
>>> constant (i.e. volume), but the values fluctuate alot (i.e.
>>> decreasing). If I am using 'switch', from say 0.8-0.9,
>>> shouldn't the values smooth off after 0.9nm? and stay
>>> constant after that? Any insights would be helpful.
>>>
>>> Thanks.
>>>
>>> Nisha P
>>>
>>>
>>> --gmx-users mailing list    gmx-users at gromacs.org
>>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>>> Please search the archive at http://www.gromacs.org/search before posting!
>>> Please don't post (un)subscribe requests to the list. Use
>>> thewww interface or send it to gmx-users-request at gromacs.org.
>>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>>
>>
>>
>>
>> -- 
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>> Please search the archive at http://www.gromacs.org/search before posting!
>> Please don't post (un)subscribe requests to the list. Use the
>> www interface or send it to gmx-users-request at gromacs.org.
>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>
>
>
>
> Nisha Patel
> MSc Candidate
> Leslie Dan Faculty of Pharmacy
> Department of Pharmaceutical Sciences
> 144 College Street
> Room 1172
> Toronto, ON
> M5S 3M2
> Canada
> Telephone: 416-978-1536
>

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar