of course if we do something, each of us hope it should be something
reliable not just can or can not.... I also looking forward the update
function for membrane protein in Amber. In fact, most people would like
to use CHARMM36 for membrane system and the problem for many tools is
the topology file for ligand. Probably, Amber developer should also
consider how to solve this problem. Since amber get Antechamber tool for
ligand building, probably someone should provide how to build CHARMM FF
based topology for ligand.
It is said that GAFF is also good enough for lipids simulation and some
paper was already published based on this. However, what I found is that
most people would like to use POPC lipids. But someone else prefer POPE
or else. Probably, Amber developer can also consider merge the POPC,POPE
parameters into the latest Amber99SB FF which can make the case much
more convenient.
Moreover, most GPCR simulation are based on homology models. So we must
be very careful about the methods used for such kind of simulation. It
would be a very good idea to introduce similar strategies in Desmond
i.e. the whole system can be relaxed very steadily and here is an
example for relaxation in Desmond:
Stage 1 - minimize
Stage 2-"Heating from 0 -> 310.0 K, H2O Barrier and Gradual restraining"
Stage 3-NPT equilibration with H2O Barrier with heavy atoms restrained
Stage 4-NPT equilibration of solvent and lipids
Stage 5-NPT with protein heavy atoms anealing 10.0 -> 2.0 kal/mol
Stage 6-NPT with C_alpha atoms restrained at 2 kcal/mol
Stage 7-NPT, with no restraints
Stage 8-MD production
The most promising steps for the above protocol are stage 2 and stage 5
because the system are restrained or unrestrained linearly which is a
perfect methods for homology based MD simulations.
Probably, Amber can also consider cooperate with Desmond and the system
builder tool is also very good in Desmond. However, what I found is that
the lipids naming methodology in Desmond is different from original
CHARMM FF if someone would like to use CHARMM36. And Amber itself
doesn't provide lipids topology which can be good enough for membrane
simulation.
On 12/21/2011 01:03 AM, Carlos Simmerling wrote:
> I'm not sure it's that simple- in principle, Amber can do membrane
> simulations and examples have been published. If one is thinking of a
> particular simulation method that is wanted to enforce aspects of membrane
> structure and perhaps stop the inaccurate force fields from doing bad
> things, then it's a different question than just whether Amber can actually
> run membrane simulations...
>
> On Tue, Dec 20, 2011 at 6:54 PM, Albert<mailmd2011.gmail.com> wrote:
>
>> As far as I know, Amber doesn't support membrane protein simulation. It
>> is said that the next version of Amber may update this function.
>>
>>
>>
>> On 12/20/2011 04:56 PM, Rajesh Singh wrote:
>>> Dear All,
>>>
>>> I am facing problems in performing MD of membrane proteins (GPCRs). I
>> would
>>> highly appreciate if anybody can provide me a tutorial for performing
>>> membrane dynamics using AMBER.
>>>
>>> Thanks and regards
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Dec 20 2011 - 16:30:04 PST
