From: Jeff Wereszczynski (jmweresz_at_mccammon.ucsd.edu)
Date: Thu Jul 07 2011 - 18:27:55 CDT
Francesco,
You've actually asked what is a difficult question that contains
many complexities. I'd suggest checking out a review on the subject before
diving in, for example Gilson and Zhou had a very nice one, here is the
reference:
Annu. Rev. Biophys. Biomol. Struct 2007. 36:21–4
Best,
Jeff
On Thu, Jul 7, 2011 at 4:29 PM, Gianluca Interlandi <
gianluca_at_u.washington.edu> wrote:
> Of course you can, but I think that Francesco is asking about an analysis
> method of his already performed trajectories.
>
> Gianluca
>
>
> On Thu, 7 Jul 2011, Jeffrey Potoff wrote:
>
> Why not use the adaptive force bias method?
>>
>> http://www.edam.uhp-nancy.fr/**ABF/ <http://www.edam.uhp-nancy.fr/ABF/>
>> http://www.ks.uiuc.edu/**Training/Tutorials/namd/ABF/**tutorial-abf.pdf
>>
>> On 7/7/2011 6:08 PM, Gianluca Interlandi wrote:
>>
>>> Francesco,
>>>
>>> Just a hint. Don't take my word that this works. I thought that free
>>> energy makes sense to calculate only when you compare it to a reference
>>> point, e.g., you have a mutation and you compare to the wild-type. So, in
>>> this way you have a DeltaG. DeltaG can be written as:
>>>
>>> DeltaG = DeltaU - T*DeltaS
>>>
>>> DeltaU is the internal energy which can be calculated with namdenergy as
>>> the interaction energy between the ligand and the protein. You want to
>>> include bulk in the calculation (although you might still want to calculate
>>> it for both, with and without bulk).
>>>
>>> The hard part, as usual, is estimating DeltaS. Schlitter had come up with
>>> a method to estimate the vibrational entropy from the covariance, which was
>>> then refined by van Gunsteren and even implemented into CHARMM (see
>>> corman.doc and grep "Schlitter"). That might give you an estimate for
>>> DeltaS. However, you need to make sure that your simulations have converged
>>> and you have sampled enough.
>>>
>>> Tutto chiaro?
>>>
>>> Gianluca
>>>
>>> On Thu, 7 Jul 2011, Francesco Pietra wrote:
>>>
>>> First, I have the systems equilibrated with charmm ff and it would be
>>>> too much work to change to amber ff. Second, the systems have lipids
>>>> and amber ffs are not so good for lipids. Third, the system is based
>>>> on a multimer protein and amber renumbers continuously all atoms,
>>>> unlike charmm that preserves subunits. Examining interactions with
>>>> continuous numbering will be headache. But it was kind from you to
>>>> give details. I wish you success with the method.
>>>> chiendarret
>>>>
>>>> On Thu, Jul 7, 2011 at 5:50 PM, Dong Luo <us917_at_yahoo.com> wrote:
>>>>
>>>>> AmberTools has utilities to calculate free energy of binding from a
>>>>> single
>>>>> MD trajectories that simulated with Amber force field. NAMD supports
>>>>> Amber
>>>>> force field.
>>>>> I recently did a test with this. It works though I am not sure of the
>>>>> quality of the results because it's the first time for me to calculate
>>>>> the
>>>>> binding free energy.
>>>>> Briefly the steps are:
>>>>> 1. Convert pdb file to Amber friendly format according to AmberTools'
>>>>> manual. The online tool:
>>>>> http://glycam.ccrc.uga.edu/**ccrc/GlycamLITE/Protein/**
>>>>> uploadIndex.jsp?option=ff99<http://glycam.ccrc.uga.edu/ccrc/GlycamLITE/Protein/uploadIndex.jsp?option=ff99>can help with it.
>>>>> 2. Using AmberTools' tleap to create parameter and coordinates input
>>>>> files
>>>>> for MD. Again check with the manual.
>>>>> 3. Edit NAMD configuration file to use the Amber force field. Check
>>>>> NAMD's
>>>>> guide.
>>>>> 4. Run the simulation.
>>>>> 5. Calculate the free energy of binding following corresponding steps
>>>>> in the
>>>>> Amber tutorial:
>>>>> http://ambermd.org/tutorials/**advanced/tutorial3/
>>>>> Dong
>>>>>
>>>>> ______________________________**__
>>>>> From: Francesco Pietra <chiendarret_at_gmail.com>
>>>>> To: NAMD <namd-l_at_ks.uiuc.edu>
>>>>> Sent: Thursday, July 7, 2011 5:52 AM
>>>>> Subject: namd-l: free energy of binding
>>>>>
>>>>> How could free energy of binding of a small-molecule ligand to a
>>>>> protein receptor be calculated from namd md trajectories?
>>>>> Is there a validated specific procedure for namd, or should a
>>>>> literature method be imitated? For example Åqvist's method:
>>>>>
>>>>> Åqvist, J., Medina, C., and Samuelsson, J. E. (1994) A new
>>>>> method for predicting binding a?nity in computer-aided drug design.
>>>>> Protein Eng. 7, 385?391.
>>>>>
>>>>> thanks for sharing experience on this hot topic
>>>>>
>>>>> francesco pietra
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>> ------------------------------**-----------------------
>>> Gianluca Interlandi, PhD gianluca_at_u.washington.edu
>>> +1 (206) 685 4435
>>> http://artemide.bioeng.**washington.edu/
>>>
>>> Postdoc at the Department of Bioengineering
>>> at the University of Washington, Seattle WA U.S.A.
>>> ------------------------------**-----------------------
>>>
>>
>>
>> --
>> ==============================**==============================**
>> ==========
>> Jeffrey J. Potoff jpotoff_at_chem1.eng.wayne.edu
>> Associate Professor Wayne State University
>> Department of Chemical Engineering and Materials Science
>> 5050 Anthony Wayne Dr Phone:(313)577-9357 Detroit, MI
>> 48202 Fax: (313)578-5815
>> http://potoff1.eng.wayne.edu
>> ==============================**==============================**
>> ==========
>>
>>
>>
> ------------------------------**-----------------------
> Gianluca Interlandi, PhD gianluca_at_u.washington.edu
> +1 (206) 685 4435
> http://artemide.bioeng.**washington.edu/
>
> Postdoc at the Department of Bioengineering
> at the University of Washington, Seattle WA U.S.A.
> ------------------------------**-----------------------
>
-- Jeff Wereszczynski NIH Postdoctoral Fellow University of California, San Diego http://mccammon.ucsd.edu/~jwereszc
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:20:33 CST