Re: [gmx-users] Very large fluctuations in dg/dl

Hi John and David, I reply to you both. Thank you very much for your complete answers as well as for all the pointers you mentioned. I will have a look at these references. Free energy calculation is definitely a challenging issue ! Cheers, Patrick John D. Chodera a écrit : Hi Patrick, I li

Re: [gmx-users] Very large fluctuations in dg/dl

Hi Stephane, just to finish with this story, I took the value 54 determined in (Hunenberger & van Gunsteren,1998,J Chem Phys; Smith & van Gunsteren, 1994, J Chem Phys), that has been updated to 61 since then. But as stated by Berk and Lars it will hardly make any difference in the properties I

Re: [gmx-users] Very large fluctuations in dg/dl

Stéphane Téletchéa wrote: Patrick Fuchs a écrit : Hi Gromacs users, I have a few questions related to solvation free energy calculation via thermodynamic integration. #ifdef FLEX_SPC #include "flexspc.itp" #else #include "spc.itp" #endif coulombtype= generalized-reaction-fi

Re: [gmx-users] Very large fluctuations in dg/dl

From: Stéphane Téletchéa <[EMAIL PROTECTED]> Reply-To: Discussion list for GROMACS users To: Discussion list for GROMACS users Subject: Re: [gmx-users] Very large fluctuations in dg/dl Date: Wed, 09 May 2007 11:11:06 +0200 Patrick Fuchs a écrit : Hi Gromacs users, I have a few que

Re: [gmx-users] Very large fluctuations in dg/dl

Patrick Fuchs a écrit : Hi Gromacs users, I have a few questions related to solvation free energy calculation via thermodynamic integration. #ifdef FLEX_SPC #include "flexspc.itp" #else #include "spc.itp" #endif coulombtype= generalized-reaction-field rcoulomb = 1.

Re: [gmx-users] Very large fluctuations in dg/dl

Patrick, My first observation was that 1 ns seemed to be a minimum for certain lambda values (e.g. lambda=0.70). I sometimes read in literature that some authors used a few hundreds of ps, which seemed (to me) not sufficient for proper convergence. My experience is that the range of lambda=0.7

Re: [gmx-users] Very large fluctuations in dg/dl

Hi Patrick, I like your plots. They nicely demonstrate the difficulty of convergence of the estimate of . It looks like there may be some other oddities in the plot, such as switching between conformations or some other effect that has a long correlation time. In particular, at lambda =

Re: [gmx-users] Very large fluctuations in dg/dl

From: Patrick Fuchs <[EMAIL PROTECTED]> Reply-To: Discussion list for GROMACS users To: Discussion list for GROMACS users Subject: Re: [gmx-users] Very large fluctuations in dg/dl Date: Tue, 8 May 2007 16:36:35 +0200 (CEST) Hi Berk, I just wanted a confirmation because it's

Re: [gmx-users] Very large fluctuations in dg/dl

for GROMACS users To: Discussion list for GROMACS users Subject: Re: [gmx-users] Very large fluctuations in dg/dl Date: Tue, 8 May 2007 12:45:32 +0200 (CEST) Hi Berk and Tsjerk, sounds reassuring since the last version of the GROMOS force field (G53a5 and a6) has been parametrized to reproduce DGsol usin

Re: [gmx-users] Very large fluctuations in dg/dl

AIL PROTECTED]> Reply-To: Discussion list for GROMACS users To: Discussion list for GROMACS users Subject: Re: [gmx-users] Very large fluctuations in dg/dl Date: Tue, 8 May 2007 12:45:32 +0200 (CEST) Hi Berk and Tsjerk, sounds reassuring since the last version of the GROMOS force field (G53a5

Re: [gmx-users] Very large fluctuations in dg/dl

Reply-To: Discussion list for GROMACS users To: "Discussion list for GROMACS users" Subject: Re: [gmx-users] Very large fluctuations in dg/dl Date: Tue, 8 May 2007 06:39:50 +0200 Hi Berk, Just a question relating to your reply (should be a new thread maybe...). The Gromos force fiel

Re: [gmx-users] Very large fluctuations in dg/dl

Hi John, thanks a lot for your reply. Indeed, the standard deviation I presented in my previous post is the one of dg/dl samples. I was just surprised by the fact the std. dev. is always larger than the value itself (since I'm starting with FE calculation I had no expectation of what the behavi

Re: [gmx-users] Very large fluctuations in dg/dl

From: "Tsjerk Wassenaar" <[EMAIL PROTECTED]> Reply-To: Discussion list for GROMACS users To: "Discussion list for GROMACS users" Subject: Re: [gmx-users] Very large fluctuations in dg/dl Date: Tue, 8 May 2007 06:39:50 +0200 Hi Berk, Just a question relating to y

Re: [gmx-users] Very large fluctuations in dg/dl

Hi Berk, Just a question relating to your reply (should be a new thread maybe...). The Gromos force fields were parameterized (for use) with cutoff/RF. PME is of course better for energy conservation, but the inclusion of the interactions beyond the cut-off may give rise to deviant behaviour from

RE: [gmx-users] Very large fluctuations in dg/dl

From: Patrick Fuchs <[EMAIL PROTECTED]> Reply-To: Discussion list for GROMACS users To: gmx-users@gromacs.org Subject: [gmx-users] Very large fluctuations in dg/dl Date: Mon, 07 May 2007 16:41:26 +0200 Hi Gromacs users, I have a few questions related to solvation free energy calculation via