[gmx-users] Re: implicit solvent and REMD

[Wei Fu]

Thanks a lot, Anton and Xavier,

	In fact, the experiment proposed that the low-PH triggers the conformational change of my protein to be unfolded state, while the detailed mechanism is unknown. My protein has a lot of charged residues, pKa calculations indicate that my protein has very different protonation states at neutral and acidic enviroment, I have run MD simulations at both protonation states for both 15ns, while I monitored the radius of gyration, SASA, H-bonds between diffrent domains, and observed the structures of trajectories, but I even did not find the trend of unfolded state of my protein at both states.  That is why I want to increase the temperature.  I do not have enough resource to run REMD, so I will try to increase the temperatue to run conventional MD and look at what will happen? I do find some papers studied folding and unfolding of protein at high temperature by running MD, can you guide me if it is reasonable to increase the temperature to run conventional MD to study the unfolded state of my protein or to see it

Wei

************************************************************
  Wei Fu (Linda)
  Postdoctoral Research Associate
  Department of Biology and Biochemistry 
  University of Houston
  http://adrik.bchs.uh.edu/~fuwei/
**************************************************************


> Wei Fu wrote:
> 
>> Hi, I want to observe the unfolded state of a protein with 370 residues,
>>
> unfolded stateSSSSSSS !!! For sure a 370 residues protein must have a lot of
> unfolded states. I think that this information is unaccessible at the time except
> if you have thousands of processors available for a long time.

For given unfolding conditions (e.g., high T, high P, high Urea), there
is *one* unfolded state. (By definition!)

However, what I think Xavier wanted to say is this state will consist of
many conformations. And he is right again in pointing out that for a protein
one cannot hope to learn much in a reliable way about the unfolded state
conformations from an MD simulation. You can certainly unfold the protein
fast (try, e.g., T=4000K), but that unfolded state may not be relevant for
what you want to do, or you can try and do it slow, but then you may not
reach the unfolded state in any reasonable amount of time.


-- 
Groetjes,

Anton