Re: [AMBER] comparing RDF plots

From: Daniel Roe <daniel.r.roe.gmail.com>
Date: Wed, 9 Apr 2014 13:20:15 -0600

On Wed, Apr 9, 2014 at 11:15 AM, newamber list <newamberlist.gmail.com> wrote:
> Actually my concern was that Avg. density increases when we use different
> masks but actually the number of waters also increases as summed for all
> shells (so numerator also increases! just not the denominator) as shown
> below. Please correct me if wrong.

It's not the number of waters that increases, but the number of
potential distances (maybe that is what you meant to say?). The
normalization factor will of course be related to the number of
potential distances you can calculate, so if you have more potential
distances there will be a correspondingly larger normalization factor.

>>>They may be qualitatively similar, but you'll have better statistics with
> the first one (and probably even better if you just make the second mask
> ':Cl-').
>
> In second case it was just one ion actually I have many such single ions in
> some region so will average for all ions and over all trajectory.

The fact that you only have one chloride in your second mask is why I
said you will have better statistics with your first mask. Of course
if all you are interested in is that particular chloride then there is
no issue.

> Summary: So in such case I think one can compare two plots quantitative as
> well. Just considering the heights of peaks should tell me whether ions in
> region of interest are more/less hydrated (second case) compared to ions in
> bulk (first case). Please suggest if it does not make sense?

It will, but just for that particular chloride ion versus chloride
ions in bulk. Also remember that cpptraj distance-based masks do not
update on a frame-by-frame basis, so if you're using distance-based
masks you have to make sure the ions are approximately where they
should be over the course of the simulation (e.g. the one that starts
out near your solute does not later move out into bulk).

-Dan

>
> best regards
> Jiomm
>
>
> On Wed, Apr 9, 2014 at 3:22 PM, Daniel Roe <daniel.r.roe.gmail.com> wrote:
>
>> Hi,
>>
>> You have a different average density because your second mask has a
>> different number of atoms, so the number of things you expect in a
>> volume slice will necessarily be different (you can see this by the
>> vastly different number of 'distances' reported). Each plot is
>> measuring different things. The first is calculating the RDF from
>> every water atom to all chloride ions within 50 Ang. of residues 484
>> and 441. The second is calculating the RDF from every water to a
>> single chloride ion. They may be qualitatively similar, but you'll
>> have better statistics with the first one (and probably even better if
>> you just make the second mask ':Cl-').
>>
>> -Dan
>>
>> On Tue, Apr 8, 2014 at 8:12 PM, Jio M <jiomm.yahoo.com> wrote:
>> > Hi All
>> >
>> > I have generated two RDF plots with different masks (for a single frame)
>> and the output shows that both are using different avg. density for rdf
>> calculation.
>> > Please correct me if I am wrong how RDF is calculated here:
>> >
>> > RDF= density*[ Vol(R+dR) - Vol(R)]
>> > Avg. density = (#mask1*#mask2)/avg. volume.
>> >
>> > I am just wondering in this case how we can compare both plots because
>> avg. density used is different in both the cases? I have pasted here both
>> commands with main outputs.
>> >
>> > 1) radial dat1 0.1 10 :WAT (:484,441>.50&:Cl-) volume
>> >
>> > ACTION OUTPUT:
>> > RADIAL: 1 frames, 502721 distances.
>> > # in mask1= 230148, # in mask2 = 1015, # in common = 0
>> > Average volume is 2555636.055524 Ang^3.
>> > Average density is 91.405902 distances / Ang^3.
>> >
>> >
>> > 2) radial dat2 0.1 10 :WAT :514 volume # (514 is an Cl- ion)
>> >
>> > ACTION OUTPUT:
>> > RADIAL: 1 frames, 396 distances.
>> > # in mask1= 230148, # in mask2 = 1, # in common = 0
>> > Average volume is 2555636.055524 Ang^3.
>> > Average density is 0.090055 distances / Ang^3.
>> >
>> > thanks
>> > Jiom
>> > _______________________________________________
>> > AMBER mailing list
>> > AMBER.ambermd.org
>> > http://lists.ambermd.org/mailman/listinfo/amber
>>
>>
>>
>> --
>> -------------------------
>> Daniel R. Roe, PhD
>> Department of Medicinal Chemistry
>> University of Utah
>> 30 South 2000 East, Room 201
>> Salt Lake City, UT 84112-5820
>> http://home.chpc.utah.edu/~cheatham/
>> (801) 587-9652
>> (801) 585-6208 (Fax)
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber



-- 
-------------------------
Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 201
Salt Lake City, UT 84112-5820
http://home.chpc.utah.edu/~cheatham/
(801) 587-9652
(801) 585-6208 (Fax)
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Apr 09 2014 - 12:30:02 PDT
Custom Search