Re: [AMBER] Cut-off distance "in Closest" command of ptraj

From: Amor San Juan <amorsanjuan.yahoo.com>
Date: Wed, 1 Sep 2010 19:41:07 -0700 (PDT)

For the benefit of others who would read this post in the future, the closest command in ptraj is not the suitable program to perform the task of filtering closest distance of water O atom to the ligand/protein. Better to use VMD for this purpose. Then feed this as input for naccess to get buried waters.

Amor

--- On Thu, 19/8/10, Jason Swails <jason.swails.gmail.com> wrote:

From: Jason Swails <jason.swails.gmail.com>
Subject: Re: [AMBER] Cut-off distance "in Closest" command of ptraj
To: "AMBER Mailing List" <amber.ambermd.org>
Date: Thursday, 19 August, 2010, 10:04 PM

I believe for each frame it calculates the distance between the COM of the
ligand (or center of geometry, not sure which one) and most likely the
oxygen atom of the water molecule (the COM of the water molecule is not much
different).  It compares the distances for all of the water molecules and
chooses the 10 that have the smallest distance.  Note, these are NOT the
same water molecules in every frame.

Good luck!
Jason

On Thu, Aug 19, 2010 at 3:14 AM, Amor San Juan <amorsanjuan.yahoo.com>wrote:

> Jason, thanks for the reply. To picture out the closest analysis, given the
> example of 50 snapshots in trajectory I issued the closest command to search
> for the 10 closest water near the ligand. The output is P+L+W10
> (protein+ligand+ten water). My inquiry is how does closest algorithm chooses
> which water are close or far? What is the distance it takes between the
> ligand and water for the algorithm to differentiate closest and far water
> molecules. TIP3P water models bounced front and back during the dynamics,
> and there are water molecules that frequently visit the hydration which
> could be long/short residency of interaction. In this case, how does closest
> algorithm choose those 10 water closest to the ligand.
>
> Amor
>
>
> --- On Wed, 18/8/10, Jason Swails <jason.swails.gmail.com> wrote:
>
> From: Jason Swails <jason.swails.gmail.com>
> Subject: Re: [AMBER] Cut-off distance "in Closest" command of ptraj
> To: "AMBER Mailing List" <amber.ambermd.org>
> Date: Wednesday, 18 August, 2010, 9:43 PM
>
> There is no cutoff.  You provide a number, and it takes that many waters
> that are closest to the selection.  If you choose 10, and the 10th closest
> water is 100 angstroms away, then the cutoff is 100 angstroms.  Am I
> misunderstanding your question?
>
> All the best,
> Jason
>
> On Wed, Aug 18, 2010 at 1:35 AM, Amor San Juan <amorsanjuan.yahoo.com
> >wrote:
>
> > What is the cut-off distance between "solute-solvent" implemented in
> > closest command? It was not specify at the manual. Is it a cut-off
> distance
> > of up to 3Angstroms? or perhaps smaller than that, am not sure.
> >
> > Thanks Jason for the previous reply, and yes I resolved the problem by
> > fixing the input script for closest in ptraj. The output trajectory
> contains
> > the first 10 closest water, followed by farthest of thousands of tip3p
> model
> > water.
> >
> > Amor
> >
> >
> >
> > --- On Tue, 17/8/10, Jason Swails <jason.swails.gmail.com> wrote:
> >
> > From: Jason Swails <jason.swails.gmail.com>
> > Subject: Re: [AMBER] Creating new topology file from close.mdcrd file
> > To: "AMBER Mailing List" <amber.ambermd.org>
> > Date: Tuesday, 17 August, 2010, 8:52 PM
> >
> > Hello,
> >
> > I'm assuming from the next email you sent out that this has been
> resolved?
> > I'm also quite certain that the output trajectory written will only have
> > the
> > 10 appropriate water molecules, not all of them with the first 10
> > corresponding to the 10 water molecules of interest.
> >
> > All the best,
> > Jason
> >
> > On Tue, Aug 17, 2010 at 5:42 AM, Amor San Juan <amorsanjuan.yahoo.com
> > >wrote:
> >
> > > Thanks Jason for always giving time and attention.
> > >
> > > The input script I used for closest is:
> > >
> > > ----
> > > trajin CLOin.mdcrd
> > > trajout CLOout.mdcrd
> > >
> > > center :1-200 mass origin
> > > image origin center
> > > solvent byres :WAT
> > > closest 10 :200 first
> > > -----
> > >
> > > The ligand numbering is 200, where I want 10 closest water around it. I
> > got
> > > the output trajectory but I bumped into another problem of generating
> the
> > > new topology file corresponding to "CLOout.mdcrd". This new topology
> file
> > > shall be obtained from a new pdb file (P+L+W10) where W10 is 10 water
> > > molecules. I wanted to do the ff steps:
> > >
> > > 1.Take one snapshot from "CLOout.mdcrd" by extraction of trajectory
> using
> > > mmpbsa.pl. Convert snapshot to pdb using ambpdb.
> > > 2. Modify the pdb file such that only 10 water molecules remained,
> > together
> > > with prot+lig
> > > 3.Create new topology using the pdb generated in step2.
> > >
> > > File CLOin.mdcrd input trajectory has 37581879 bytes while CLOout.mdcrd
> > > output trajectory has only 4161681 bytes. Input traj contains 9000
> TIP3P
> > > model waters, and after executing closest command the output is
> expected
> > to
> > > contain only 10 closest water molecules (but Im not sure of this
> > > assumption).
> > >
> > > My question is: does the output trajectory from "closest" command gives
> > > only the specified number of water? Or, the selected 10 closest water
> are
> > > written immediately after prot+lig, and followed by the rest of
> thousands
> > of
> > > tip3p models which are farthest. I need to know precisely so that I
> will
> > > track down the error of generating snaphots from CLOout.mdcrd.
> > >
> > > In this case, the CLOin.mdcrd has (Prot+Lig+Wat)=PLW.prmtop contains
> > 30324
> > > atoms. This PLW.prmtop does not correspond to CLOout.mdcrd, because
> > whenever
> > > I tried generating snapshots there is mismatch of atoms. I coudnt be
> able
> > to
> > > generate a pdb file from step1 above.
> > >
> > > Mismatch atoms in generating pdb seems to me that PLW10.prmtop should
> be
> > > used, but to where I could get it as this is the reason I am doing
> steps
> > 1-3
> > > to get it.
> > >
> > > Please anybody help me get from this bump. I have exhausted the entire
> > day.
> > >
> > > Amor
> > >
> > >
> > >
> > > --- On Mon, 16/8/10, Jason Swails <jason.swails.gmail.com> wrote:
> > >
> > > From: Jason Swails <jason.swails.gmail.com>
> > > Subject: Re: [AMBER] How to identify water molecules at the interface ?
> > > follow up ....
> > > To: "AMBER Mailing List" <amber.ambermd.org>
> > > Date: Monday, 16 August, 2010, 9:03 PM
> > >
> > > Hello,
> > >
> > > On Mon, Aug 16, 2010 at 2:19 AM, Amor San Juan <amorsanjuan.yahoo.com
> > > >wrote:
> > >
> > > > I want to write a follow-up in regard to my previous post below.
> > > >
> > > > Aside from ptraj hbond analysis, I also tried several ptraj commands
> > > > including:
> > > > a) grid = the aim of this command is to generate water density at
> > > hydration
> > > > b)watershell= the aim of this command is to count the number of water
> > > > molecules at 1st/2nd solvation shell
> > > > c) closest= the aim of this command is to quantify the number of
> water
> > > > molecules at 3.5A and 5A.
> > > >
> > > > In all those commands above, I am unable to identify the numbering of
> > > water
> > > > buried at interface. Although the closest command can quantify the
> > water
> > > > molecules at specific distance, its numbering of water molecules were
> > > > altered compared to the original numbering.
> > > >
> > > > My aim is to get the identity of the water molecules buried between
> the
> > > > ligand-protein. Once the water identity is found, I will use this as
> > > input
> > > > together with prot+lig for binding energy calculation in mmpbsa. What
> I
> > > mean
> > > > to say in water identity is simply the water identity as reflected on
> > its
> > > > atom/residue number.
> > > >
> > >
> > > You don't need this really.  Just create a new prmtop and trajectory
> file
> > > with only the proper number of waters that you kept with the "closest"
> > > command.  Then it doesn't matter what the original numbering was in the
> > > prmtop/trajectory file(s), it's simply the new numbers it has in the
> new
> > > trajectory.  Then feed this into MMPBSA and don't tell it to strip any
> > > waters/ions.  For example, if you're using MMPBSA.py, set strip_mdcrd=0
> > to
> > > avoid stripping the trajectory file of any waters/ions.  For
> mm_pbsa.pl,
> > > just include the water atoms as part of the ligand or receptor atoms.
> > >
> > > Hope this helps,
> > > Jason
> > >
> > >
> > > > amor
> > > >
> > > >
> > > > --- On Mon, 16/8/10, Amor San Juan <amorsanjuan.yahoo.com> wrote:
> > > >
> > > > From: Amor San Juan <amorsanjuan.yahoo.com>
> > > > Subject: How to identify water molecules at the interface ?
> > > > To: "AMBER Mailing List" <amber.ambermd.org>
> > > > Date: Monday, 16 August, 2010, 11:51 AM
> > > >
> > > > I have read several posts regarding water calculations in several
> ways
> > > for
> > > > different purposes done. But it seems to me that the goal I want to
> > > achieve
> > > > needs relevant feedback from the forum.
> > > >
> > > > I have a system (protein + peptide + ligand) with 200 residues in
> total
> > > and
> > > > note that ligand has a phosphate group. I want to identify the water
> > > > molecules (TIP3P models) found at the interface between the ligand
> and
> > > > protein. I need to identify which water molecules are at the
> interface
> > so
> > > I
> > > > can use it as input for binding energy calculation by MMPBSA. Here
> are
> > > the
> > > > steps formulated for water identification:
> > > >
> > > > 1) Calculate the water within 3.5Angstroms using the trajectory
> input.
> > > The
> > > > expected outcome is (P+L+Wi) such that P=protein, L=ligand, and
> > Wi=water
> > > at
> > > > ith frame. From the output, remove all hydrogen molecules in water.
> > > >
> > > > 2)Calculate accessibility by
> > > >  filtering out WAT.O (Water oxygen) within >0.50Angstroms
> > > >
> > > > 3) Repeat step 2 to get only the buried water molecules remaining at
> > > > interface.
> > > >
> > > > One suggestion I got is to make a script to do in one shot the above
> > > three
> > > > steps. I think about it but am not sure if one straightforward script
> > can
> > > do
> > > > the task.
> > > >
> > > > I approached the problem by trying to do it one-step-at-a-time. For
> > step
> > > 1
> > > > above, I implemented ptraj. Using ptraj h-bond distance 3.5 analysis,
> > it
> > > > gave me an output of P+L+Wi, however looking at Wi seems to be so
> > > enormous
> > > > from the list as expected.
> > > > For step2, I need the trajectory that contains P+L+Wi with stripped
> > > > hydrogens generated from step1. I have not proceeded to step2 yet.
> > > >
> > > > To get the trajectory P+L+Wi, it seems the way out is to use the
> > > > ptraj-hbond output to select interacting atoms, but as noted Wi in
> the
> > > list
> > > > is quite long to do manual selection one-by-one.
> > > >
> > > >  I also thought of using LIGPLOT program to
> > > >  identify buried water molecules but my input is a trajectory not a
> > pdb.
> > > I
> > > > know I can generate pdb from snapshots in trajectory but how would
> you
> > > take
> > > > it for several thousands of snapshots pdb? Bottomline, I only need to
> > > > identify the water at interface based on the trajectory.
> > > >
> > > > Amor
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > _______________________________________________
> > > > AMBER mailing list
> > > > AMBER.ambermd.org
> > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > >
> > >
> > >
> > >
> > > --
> > > Jason M. Swails
> > > Quantum Theory Project,
> > > University of Florida
> > > Ph.D. Graduate Student
> > > 352-392-4032
> > > _______________________________________________
> > > 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
> > >
> >
> >
> >
> > --
> > Jason M. Swails
> > Quantum Theory Project,
> > University of Florida
> > Ph.D. Graduate Student
> > 352-392-4032
> > _______________________________________________
> > 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
> >
>
>
>
> --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Graduate Student
> 352-392-4032
> _______________________________________________
> 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
>



-- 
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
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
Received on Wed Sep 01 2010 - 20:00:03 PDT
Custom Search