RE: Help in reverse Free Energy perturbation - hysteresis

From: Bruno Luís Pinto de Oliveira (boliveira_at_itn.pt)
Date: Tue Oct 04 2011 - 14:29:34 CDT

Thanks Chris and Jerome for your reply,

> The data you are plotting is relative free energy: it is defined up to
> an additive constant. The best way to represent the change along the
> forward-and-backward simulation is to shift the backward curve by a
> constant, so that the initial value (at lambda = 1) is the same free
> energy as the final value from the forward run.

I corrected the backward curve by subtracting the free energy obtained for lambda =1 in the foward calculation.

Please find the graphic published here:

http://www.boliveira.com/NAMD.html

I'm not familiar enough with the FEP to see if the result is acceptable...

I guess it is not so bad ...

or Do I have to do longer simulations ???

I tried to analyse the FEP calculations with the ParseFEP plugin as suggested by Chris Chipot however I am getting this error:

domain error: argument not in valid range
domain error: argument not in valid range
while executing
"expr sqrt($instant_error_gauss) "
(procedure "::ParseFEP::FEP_formula" line 361)
invoked from within
"::ParseFEP::FEP_formula $file $file_entropy $file_lambda $window $mean_xi $fororback"
(procedure "::ParseFEP::normal_parse_log" line 60)
invoked from within
"::ParseFEP::normal_parse_log $::ParseFEP::fepofile forward "
(procedure "::ParseFEP::namdparse" line 172)
invoked from within
"::ParseFEP::namdparse"
(in namespace inscope "::ParseFEP" script line 21)
invoked from within
"::namespace inscope ::ParseFEP {
##if {$::ParseFEP::fepbofile != ""} {set ::ParseFEP::sosindex 1 }

if { [string length $fepofile] < 1 } {
tk_dial..."
invoked from within
".parseFEP.runbutton invoke"
("uplevel" body line 1)
invoked from within
"uplevel #0 [list $w invoke]"
(procedure "tk::ButtonUp" line 24)
invoked from within
"tk::ButtonUp .parseFEP.runbutton"
(command bound to event)

I found this error reported in a previous message posted in the mailling list.

http://www.ks.uiuc.edu/Research/vmd/mailing_list/vmd-l/17776.html

It seems that this error is related with some bug in parseFEP...

However it is strange because I have instaled the last version of VMD
(VMD 1.9 Windows OpenGL, CUDA (Windows XP/Vista/7 (32-bit) with OpenGL and CUDA).

Can someone confirm if this error is related with some bug or if I am doing something wrong???

Tkanks you in advance

Bruno

>> Dear all,
>>
>> I am new to FEP and since the last months I have gotten great help from the
>> mailing list. Thanks to all who help me...
>>
>> I am posting this message because I have a new doubt about the reverse
>> calculation to
>> monitor the hysteresis.
>>
>> Briefly, I would like to estimate the difference in the binding energy
>> between a Rhenium complex and the corresponding ligand (compounds are
>> exactly the same with the exception that the ligand lacks the metallic part
>> Re(CO)3; the Re(CO)3 vanish).
>>
>> I did the foward transformation (transformation in water and protein; lambda
>> goes from 0 to 1) and the calculated results are in good agreement with the
>> experimental results (I used a thermodynamic cycle that someone in a
>> previous post confirmed that is correct; I am using 16 ps of equilibration
>> and 80 ps of colection per lambda, 32 states, ~3ns).
>>
>> Then, I used the end of the foward transformation as the starting point for
>> the reverse simulation (lambda goes from 1 to 0).
>>
>> For example, the transformation of the Rhenium complex into the ligand gave
>> -31.33 kcal/mol (forward) and 35.34 kcal/mol (backward).
>> From the literature that I read I realized that the backward transformation
>> should display more or less the same free energy change of the forward
>> transformation but with an opposite sign. Did I understood well?
>>
>> if so can I assume that the hysteresis is acceptable???
>>
>> or I need to plot the DG against the lambda as examplified in the tutorial
>> of Christophe Chipot et al.
>>
>> I did it and as you can see in the graphic the foward and backward curves do
>> not converge.
>>
>> http://www.boliveira.com/NAMD.html
>>
>> I am confuse because if the values of the backward and foward transformation
>> should have opposite sign of course they will never converge.
>>
>> I feel that I'm missing some step but I can not figure out what...
>>
>> Excuse me the long post but all my group works with amber/TI and I dont have
>> anyone who can help me.
>>
>> Can anyone help me in all these doubts??
>>
>> Any help is highly appreciated...
>>
>> Thanks in advance
>>
>> Bruno
>>
>>

--
_______________________________________________________________________
Chris Chipot, Ph.D.
Theoretical and Computational Biophysics Group
Beckman Institute
University of Illinois at Urbana-Champaign
405 North Mathews Phone: (217) 244-5711
Urbana, Illinois 61801 Fax: (217) 244-6078
E-mail: chipot_at_ks.uiuc.edu
Christophe.Chipot_at_edam.uhp-nancy.fr
Web: http://www.ks.uiuc.edu/~chipot
http://www.edam.uhp-nancy.fr
The light shines in the darkness, and the darkness has not overcome it.
John 1:5.
_______________________________________________________________________

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:20:52 CST