AMBER Archive (2008)

Subject: FW: FW: AMBER: FW: MD run time inquiry

From: Campbell, Patrick (pcampbell_at_msm.edu)
Date: Thu Mar 27 2008 - 08:55:28 CDT

 

________________________________

From: Campbell, Patrick
Sent: Thu 3/27/2008 8:54 AM
To: amber_at_scripps.edu
Subject: RE: FW: AMBER: FW: MD run time inquiry

Thanks for your quick response Dr.Simmerling.
 
I do have a homology model of the protein and would like to refine its structure to make it as accurate as possible. Ideally, I would like to see its folding pattern if this is practical.
Can you suggest a possible strategy that I could employ?
 
Thanks much.
Patrick
 
 

________________________________

From: owner-amber_at_scripps.edu on behalf of Carlos Simmerling
Sent: Thu 3/27/2008 8:15 AM
To: amber_at_scripps.edu
Subject: Re: FW: AMBER: FW: MD run time inquiry

do you mean you want to predict the structure or see how it folds?
even small proteins usually take microseconds to fold, so for something
that large it is not practical to use all atom MD to fold it. If you have
a homology model that you want to refine, that may be possible.
let us know if your goal is to fold it.
CS

On Thu, Mar 27, 2008 at 8:14 AM, Campbell, Patrick <pcampbell_at_msm.edu> wrote:

         

________________________________

        From: Campbell, Patrick
        Sent: Thu 3/27/2008 7:06 AM
        To: amber_at_scripps.edu
        Subject: RE: AMBER: FW: MD run time inquiry
        
        
        Thanks very much for your input Dr. Simmerling!
        I would like to examine the folding pattern of a 300 aa protein as achieved at the end of an MD simulation
        (explicit solvent). While ideally I would like to have a population of alternate structures, I would say the
        limiting factor in this regard will be the time allocation (nanoseconds) for this objective. Again, as you
        mentioned, reproducibility would be the key in validating the subsequent results and following on this, the experiment
        will be repeated until such information is achieved.
        My question then is this. Based on your previous response and the information indicated above, will a time frame of
        5 nanoseconds be sufficient for the MD simulation?
        I do hope I have been clear in the representation of my question but if I am not, do let me know what additional
        information will be required to add further clarity to me inquiry.
        Thanks again and do have a great day!
        Patrick
________________________________

        From: owner-amber_at_scripps.edu on behalf of Carlos Simmerling
        Sent: Thu 3/27/2008 5:41 AM
        To: amber_at_scripps.edu
        Subject: Re: AMBER: FW: MD run time inquiry
        
        
        it depends very much on what you want to study and what actually
        happens. A good rule of thumb is that whatever happens, it should be reproducible.
        Also you need to know something about different timescales of motion- if you want
        to look at folding, then clearly 1ns is not enough. if you want to look at side chain
        motion, it may be. Another point is whether you want your data to give a hint
        of what might happen, or if you want to calculate actual populations of alternate
        structures. The latter probably requires observation of multiple transitions.
        If your goal is to study loop dynamics as in the example, then I would say run it
        until whatever you are measuring or reporting gives the same results in two independent
        runs. I would expect loop motions to take multiple ns at minimum, but some loops
        move more slowly than others so if you don't have experimental data you will just need
        to run it out and see.
        
        
        On Thu, Mar 27, 2008 at 6:13 AM, Campbell, Patrick <pcampbell_at_msm.edu> wrote:
        

                 

________________________________

                From: Campbell, Patrick
                Sent: Wed 3/26/2008 10:18 PM
                To: amber_at_scripps.edu
                Subject: MD run time inquiry
                
                
                A pleasant good day to you all,
                 
                I am currently doing an MD simulation, following on the example of Dr. Matt Lee (Loop dynamics of the HIV-1 integrase core domain) on the AMBER scripps site. Thus far, I have done a simulation for 300 picoseconds and will be looking to extend this time to 1.1 nanoseconds. My question is this - is there an established amount of time that would be deemed as sufficient for this simulation to run, such that the results will be acceptable for publication?
                 
                Thanks again for all your assistance and do have a great day!
                Patrick
                
                 
                 

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu