Hi Callum,
Thanks for your answer! I suppose that csurften=0, gamma_ten=0 is equal to csurften=3, gamma_ten=0?
Regards,
Filip
________________________________
From: "Dickson, Callum J" <callum.dickson09.imperial.ac.uk>
To: filip fratev <filipfratev.yahoo.com>
Sent: Friday, March 21, 2014 12:15 AM
Subject: RE: [AMBER] PgP protein and amber14 lipid FF
Hi Filip,
I was not able to find the figures you attached to your initial email, could you send them to me?
I have never tried a protein as large as PgP so I honestly am not sure what will happen. I have tried a few smaller proteins with anisotropic NPT (ntp=2), you could also try semiistropic but no surface tension (ntp=3, csurften=0, gamma_ten=0, ninterface=2) as many people use this in the literature.
When I first saw your email I wasn't sure what to suggest, possibly a larger bilayer as it sounds like there isn't a great deal of lipid separating the protein and the edge of the box.
All the best,
Callum
________________________________________
From: amber-bounces.ambermd.org [amber-bounces.ambermd.org] on behalf of filip fratev [filipfratev.yahoo.com]
Sent: 17 March 2014 11:10
To: AMBER Mailing List
Subject: Re: [AMBER] PgP protein and amber14 lipid FF
Hi,
I saw that someone here tried with gamma_ten=0 (just lipids) as should be but personally I was not able to perform any adequate calculations with Lipid 14 FF, in protein environment and POPC without gamma_ten=17. Just share my experience..:)
Regards,
Filip
On Friday, March 7, 2014 7:14 AM, filip fratev <filipfratev.yahoo.com> wrote:
Or I should just use csurften=3, ninterface=2, gamma_ten=17 (POPC) ?
Regards,
Filip
On Friday, March 7, 2014 2:36 AM, filip fratev <filipfratev.yahoo.com> wrote:
Hi,
I am new to the membrane simulations and have
few probably trivial questions.
I wish to simulate the PgP protein and the corresponding
lipid bilayer was created by Charmm-gui. Unfortunately in
Charmm-gui the system size along the X and Y must be the same but the
PgP is asymmetric and has dimensions about 90, 50, and 120. The initial box that
was created by Charmm-gui was 108x108x177 not as I wished - X=100, Y=130 A.
Further, I use the new Amber lipid 14 FF and
the script that has been supplied for membrane equilibration:
http://archive.ambermd.org/201401/0397.html
http://archive.ambermd.org/201401/0379.html
The only difference was the cutoff of 8.0A, gamma_lnwas set to 2.0and I use iwrap=-1. The protein was
present but harmonically restricted. I set the box by setBox Unit vdw command
and the initial box size was X=134.7, Y=138 and Z=179 A. During the 10th box relaxing steps it decreased significantly and I started the production,
i.e. membrane equilibration with box size of 106x107x174A. The equilibration
was set to 20ns, which I know is small but the size of the system is 200K
atoms. Calculations were performed by GPU Amber 12 version. It is difficult to
calculate the size/lipid due to the protein pore but roughly it was about 73
after box relaxing and decreased to 69 after 20ns which seems to be good. However,
as you can see from the attached pictures (before and after 20ns of equilibration)
the membrane curved significantly. I suppose that this is due to the protein
asymmetry introduced to the membrane symmetry but it might be everything…FF,
input parameters, etc…On the other hand why the membrane should be planar in
reality? I mean that if the box was 300x300A I would probably obtain the same asymmetry?
Is it reasonable to continue with the production run with protein unrestricted?
What is your opinion about this
simulation/situation :)?
Regards,
Filip
_______________________________________________
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 Thu Mar 20 2014 - 16:30:03 PDT
