Re: FEP calculations: optimizing the dlambda, number of steps

From: Shyno Mathew (sm3334_at_columbia.edu)
Date: Fri Nov 22 2013 - 10:32:44 CST

Thanks for the reply Jerome.
Now, for each window I have 5000 minimization followed by 10000 steps. I
did try going to 10^6 steps, but didn't give significant difference in free
energy.
So just to be clear, in the previous message by saying "increase the
sampling time for just those two by running extra simulations" you meant
going to smaller dlambda values between 0.45 and 0.55?
Also, does it make sense to do FEP just between 0.45 and 0.55 and then use
it with data for all other windows?
Or should I do all this as part of one run?

thanks,
Shyno

On Fri, Nov 22, 2013 at 5:21 AM, Jérôme Hénin <jerome.henin_at_ibpc.fr> wrote:

> Hi Shyno,
>
> In general, what's available is documented in the user's guide - although
> I admit there can sometimes be a lag after new features are added.
>
> Anyway, in this particular case, the answer is no. You'll have to use
> "statically modified windows". You could also keep the current windows and
> increase the sampling time for just those two by running extra simulations.
>
> Cheers,
> Jerome
>
>
> ----- Original Message -----
> > Hey Jerome,
> > thanks again fro your reply.
> > When I plot the probability distribution for different consecutive lambda
> > windows, most plots show good overlap except the following:
> > lambda from 0.45 to 0.5 and lambda from 0.5 to 0.55.
> > So I am looking into the option of reducing dlambda for the window 0.45
> to
> > 0.55. It seems in gibbs (AMBER module for FEP) there is an option of
> doing
> > FEP with dynamically modified windows.
> > Is there anything like this in namd?
> >
> > thanks,
> > Shyno
> >
> >
> > On Tue, Nov 19, 2013 at 5:25 AM, Jérôme Hénin <jerome.henin_at_ibpc.fr>
> wrote:
> >
> > >
> > > > The column before temperature is the average change in potential
> energy
> > > > (deltaUavg) and the column preceding it is the change in potential
> energy
> > > > (delta U). Is this correct?
> > >
> > > Yes, that is correct.
> > >
> > > Jerome
> > >
> > >
> > > > On Mon, Nov 18, 2013 at 6:32 PM, Jérôme Hénin <jerome.henin_at_ibpc.fr>
> > > wrote:
> > > >
> > > > > Hi,
> > > > >
> > > > > AlchDecouple only makes a difference if there are nonbonded
> > > interactions
> > > > > within the perturbed groups.
> > > > >
> > > > > "AlchDecouple on" means that these interactions are unperturbed, so
> > > that
> > > > > the "intact" molecule is decoupled from its environment.
> > > > > "AlchDecouple off" means that the molecule's nonbonded interactions
> > > > > (included its self-interactions) are perturbed.
> > > > >
> > > > > Jerome
> > > > >
> > > > >
> > > > > ----- Original Message -----
> > > > > > also, can some one explain the parameter "AlchDecouple"
> > > > > > I have done runs with both " alchDecouple on" and "alchDecouple
> > > off", but
> > > > > > didn't notice any difference while plotting delta G vs.
> deltalambda
> > > > > values.
> > > > > >
> > > > > > Any help will be really appreciated,
> > > > > >
> > > > > > thanks,
> > > > > > Shyno
> > > > > >
> > > > > >
> > > > > > On Mon, Nov 18, 2013 at 12:44 PM, Shyno Mathew <
> sm3334_at_columbia.edu>
> > > > > wrote:
> > > > > >
> > > > > > > Hello all,
> > > > > > > I think I figured out the answers
> > > > > > > The column before temperature is the average change in
> potential
> > > energy
> > > > > > > (deltaUavg) and the column preceding it is the change in
> potential
> > > > > energy
> > > > > > > (delta U). So I should calculate the variance of delta U
> values to
> > > > > optimize
> > > > > > > delta lambda, length of each window?
> > > > > > > Could some one confirm this?
> > > > > > >
> > > > > > > thanks,
> > > > > > > Shyno
> > > > > > >
> > > > > > >
> > > > > > > On Fri, Nov 15, 2013 at 4:02 PM, Shyno Mathew <
> sm3334_at_columbia.edu
> > > >
> > > > > wrote:
> > > > > > >
> > > > > > >> Dear all,
> > > > > > >> I am doing the FEP calculations and need some help in
> > > understanding
> > > > > the
> > > > > > >> .fepout file.
> > > > > > >> 1. this is the first part of .fepout file
> > > > > > >> # STEP Elec vdW dE dE_avg Temp dG
> > > > > > >>
> > > > > > >> # l l+dl l l+dl E(l+dl)-E(l)
> > > > > > >>
> > > > > > >>
> > > > > > >> #NEW FEP WINDOW: LAMBDA SET TO 0 LAMBDA2 0.05
> > > > > > >> FepEnergy: 0 -469793 -469790 34401.03 34399.75 0.8153 0.8153
> > > 298.6998
> > > > > > >> 0.8153 FepEnergy: 100 -472878 -472876 33902.44 33900.11
> 0.4059
> > > 0.3622
> > > > > > >> 313.4741 0.1558 FepEnergy: 200 -472922 -472919 33338.32
> 33336.3
> > > > > 0.8862
> > > > > > >> 0.1122 312.7195 -0.3716
> > > > > > >> I understand the penultimate column is temperature and the
> last
> > > > > column is
> > > > > > >> dG value. What exactly are the other columns?
> > > > > > >>
> > > > > > >> 2. I am trying to see what dlambda, number of steps will give
> me
> > > > > accurate
> > > > > > >> results.
> > > > > > >> Following article says keeping the variance at each stage on
> the
> > > > > order of
> > > > > > >> 1-2 KT usually guarantees good overlap between probability
> > > > > distributions
> > > > > > >> in
> > > > > > >> two consecutive strata and a reliable error estimate.
> > > > > > >> Here the variance means the variance of dG values (only of the
> > > > > ensemble
> > > > > > >> average, not including minimization??) for a given lambda
> window?
> > > > > > >> http://pubs.acs.org/doi/abs/10.1021/jp102971x
> > > > > > >> (talking about page 10242 of article)
> > > > > > >> For 300 K, 1-2 KT is 0.596 - 1.192 Kcal/mol.
> > > > > > >> The variance (considering ensemble (not minimization values)
> dG
> > > > > values
> > > > > > >> for first window, lambd= 0 to lambda=0.05) is 0.0158. So it's
> one
> > > > > order
> > > > > > >> lower. This means I need to change dlambda and/or change the
> > > number of
> > > > > > >> steps after minimization?
> > > > > > >>
> > > > > > >> thanks,
> > > > > > >> Shyno
> > > > > > >>
> > > > > > >> --
> > > > > > >> Shyno Mathew
> > > > > > >> PhD student
> > > > > > >> Department of Chemical Engineering
> > > > > > >> Columbia University
> > > > > > >>
> > > > > > >>
> > > > > > >
> > > > > > >
> > > > > > > --
> > > > > > > Shyno Mathew
> > > > > > > PhD student
> > > > > > > Department of Chemical Engineering
> > > > > > > Columbia University
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > --
> > > > > > Shyno Mathew
> > > > > > PhD student
> > > > > > Department of Chemical Engineering
> > > > > > Columbia University
> > > > > >
> > > > >
> > > >
> > > >
> > > >
> > > > --
> > > > Shyno Mathew
> > > > PhD student
> > > > Department of Chemical Engineering
> > > > Columbia University
> > > >
> > >
> >
> >
> >
> > --
> > Shyno Mathew
> > PhD student
> > Department of Chemical Engineering
> > Columbia University
> >
>

-- 
Shyno Mathew
PhD student
Department of Chemical Engineering
Columbia University

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:56 CST