- Posts: 19
Membrane protein insertion in bilayer - mdp files
- almeida
-
Topic Author
- Offline
- Fresh Boarder
Less
More
8 years 6 months ago #4921
by almeida
Membrane protein insertion in bilayer - mdp files was created by almeida
Hi all,
I am running CGMD simulation of a membrane protein in DPPC with Martini 2.2. I am simulating my protein as standar CG mmodel and using elastic network, including ElNeDyn (as in tutorials), to compare the results. The runs goal is to to insert the protein in the membrane, starting from different orientations, keeping the secondary structure.
I just want to check if these mdp files are correct. I would appreciate if you take a look and tell me if there is something wrong or missing. Any sugestion or advice will be appreciated.
<minimization.mdp>
integrator = steep
dt = 0.02
nsteps = 500
nstxout = 0
nstvout = 0
nstlog = 100
nstxtcout = 100
xtc-precision = 10
rlist = 1.4
coulombtype = shift
rcoulomb = 1.2
epsilon_r = 15
vdw-type = shift
rvdw-switch = 0.9
rvdw = 1.2
tcoupl = v-rescale
tc-grps = Protein DPPC W_Ion
tau-t = 1.0 1.0 1.0
ref-t = 300 300 300
define = -DPOSRES
<equilibration.mdp>
dt = 0.02
nsteps = 25000
nstxout = 0
nstvout = 0
nstlog = 100
nstxtcout = 100
xtc-precision = 100
rlist = 1.4
coulombtype = shift
rcoulomb = 1.2
epsilon_r = 15
vdw-type = shift
rvdw-switch = 0.9
rvdw = 1.2
tcoupl = v-rescale
tc-grps = Protein DPPC W_Ion
tau-t = 1.0 1.0 1.0
ref-t = 300 300 300
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau-p = 12.0 12.0
compressibility = 3e-4 3e-4
ref-p = 1.0 1.0
refcoord_scaling = all
<md_2us.mdp>
dt = 0.02
nsteps = 100000000
nstxout = 0
nstvout = 0
nstlog = 10000
nstxtcout = 1000
xtc-precision = 100
rlist = 1.4
coulombtype = shift
rcoulomb = 1.2
epsilon_r = 15
vdw-type = shift
rvdw-switch = 0.9
rvdw = 1.2
tcoupl = v-rescale
tc-grps = Protein DPPC W_Ion
tau-t = 1.0 1.0 1.0
ref-t = 300 300 300
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau-p = 12.0 12.0
compressibility = 3e-4 3e-4
ref-p = 1.0 1.0
Thanks in advance
Yasser
I am running CGMD simulation of a membrane protein in DPPC with Martini 2.2. I am simulating my protein as standar CG mmodel and using elastic network, including ElNeDyn (as in tutorials), to compare the results. The runs goal is to to insert the protein in the membrane, starting from different orientations, keeping the secondary structure.
I just want to check if these mdp files are correct. I would appreciate if you take a look and tell me if there is something wrong or missing. Any sugestion or advice will be appreciated.
<minimization.mdp>
integrator = steep
dt = 0.02
nsteps = 500
nstxout = 0
nstvout = 0
nstlog = 100
nstxtcout = 100
xtc-precision = 10
rlist = 1.4
coulombtype = shift
rcoulomb = 1.2
epsilon_r = 15
vdw-type = shift
rvdw-switch = 0.9
rvdw = 1.2
tcoupl = v-rescale
tc-grps = Protein DPPC W_Ion
tau-t = 1.0 1.0 1.0
ref-t = 300 300 300
define = -DPOSRES
<equilibration.mdp>
dt = 0.02
nsteps = 25000
nstxout = 0
nstvout = 0
nstlog = 100
nstxtcout = 100
xtc-precision = 100
rlist = 1.4
coulombtype = shift
rcoulomb = 1.2
epsilon_r = 15
vdw-type = shift
rvdw-switch = 0.9
rvdw = 1.2
tcoupl = v-rescale
tc-grps = Protein DPPC W_Ion
tau-t = 1.0 1.0 1.0
ref-t = 300 300 300
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau-p = 12.0 12.0
compressibility = 3e-4 3e-4
ref-p = 1.0 1.0
refcoord_scaling = all
<md_2us.mdp>
dt = 0.02
nsteps = 100000000
nstxout = 0
nstvout = 0
nstlog = 10000
nstxtcout = 1000
xtc-precision = 100
rlist = 1.4
coulombtype = shift
rcoulomb = 1.2
epsilon_r = 15
vdw-type = shift
rvdw-switch = 0.9
rvdw = 1.2
tcoupl = v-rescale
tc-grps = Protein DPPC W_Ion
tau-t = 1.0 1.0 1.0
ref-t = 300 300 300
Pcoupl = parrinello-rahman
Pcoupltype = semiisotropic
tau-p = 12.0 12.0
compressibility = 3e-4 3e-4
ref-p = 1.0 1.0
Thanks in advance
Yasser
Please Log in or Create an account to join the conversation.
- mnmelo
-
- Offline
- Admin
Less
More
- Posts: 89
8 years 6 months ago #4990
by mnmelo
Replied by mnmelo on topic Membrane protein insertion in bilayer - mdp files
Hi Yasser,
A lot will depend on how the insertion is being carried out. Are you simply overlapping a protein with a membrane? That will probably land you in trouble unless you take special precautions. Or are you using the insane.py script?
Assuming you're doing it the insane.py way these .mdps need only a little fixing. If you're simply overlapping structures let me know because then you need a different recipe.
Some notes on your current .mdps:
Minimization
- Depending on how artificial the initial configuration is (if you're having a lot of overlaps) minimization should be run for more steps (I'd do 4000).
- Keeping position restraints during the energy minimization might prevent your system from minimizing properly (depending, of course, on exactly which restraints are in place). For such a short minimization I'd say there's little risk of significant overall displacement of system components.
- You can clean up your minimization.mdp a lot. In an energy minimization there is no time or velocities. Therefore the dt option and all the temperature coupling ones are unneeded and will be ignored. There's also no .xtc output, so xtc-related options are unneeded.
Equilibration/Production
- In your equilibration.mdp using the Parinello-Rahman pressure coupling straight from the just-minimized structure will probably cause box size to oscillate and diverge. Equilibrate with Berendsen (with a lower tau) and use Parinello-Rahman only for production.
- You are using the refcoord-scaling option for equilibration, but this only makes sense together with position restraints and there's no 'define' to indicate that you're using those. Also, refcoord-scaling=all might distort your protein if you're restraining more than one bead. refcoord-scaling=com is probably best (again, depending on the exact setup).
- You should temperature-couple the protein together with either the aqueous phase (if it sticks mostly out of the membrane) or the membrane phase (if it's mostly transmembranar). On its own it likely has too few degrees of freedom for proper weak coupling.
Good luck,
Manel
A lot will depend on how the insertion is being carried out. Are you simply overlapping a protein with a membrane? That will probably land you in trouble unless you take special precautions. Or are you using the insane.py script?
Assuming you're doing it the insane.py way these .mdps need only a little fixing. If you're simply overlapping structures let me know because then you need a different recipe.
Some notes on your current .mdps:
Minimization
- Depending on how artificial the initial configuration is (if you're having a lot of overlaps) minimization should be run for more steps (I'd do 4000).
- Keeping position restraints during the energy minimization might prevent your system from minimizing properly (depending, of course, on exactly which restraints are in place). For such a short minimization I'd say there's little risk of significant overall displacement of system components.
- You can clean up your minimization.mdp a lot. In an energy minimization there is no time or velocities. Therefore the dt option and all the temperature coupling ones are unneeded and will be ignored. There's also no .xtc output, so xtc-related options are unneeded.
Equilibration/Production
- In your equilibration.mdp using the Parinello-Rahman pressure coupling straight from the just-minimized structure will probably cause box size to oscillate and diverge. Equilibrate with Berendsen (with a lower tau) and use Parinello-Rahman only for production.
- You are using the refcoord-scaling option for equilibration, but this only makes sense together with position restraints and there's no 'define' to indicate that you're using those. Also, refcoord-scaling=all might distort your protein if you're restraining more than one bead. refcoord-scaling=com is probably best (again, depending on the exact setup).
- You should temperature-couple the protein together with either the aqueous phase (if it sticks mostly out of the membrane) or the membrane phase (if it's mostly transmembranar). On its own it likely has too few degrees of freedom for proper weak coupling.
Good luck,
Manel
Please Log in or Create an account to join the conversation.
- almeida
-
Topic Author
- Offline
- Fresh Boarder
Less
More
- Posts: 19
8 years 6 months ago #5001
by almeida
Replied by almeida on topic Membrane protein insertion in bilayer - mdp files
Hello Manel,
I am using insane.py script. Basically, I am following the Martini protein tutorials with my membrane protein. My focus is to see whether my protein is inserted in the membrane, comparing standard run with elastic network (standard EN and ElNeDyn). I am running my protein starting with coordinates where the protein is tipically inserted along the TM helices (I called "0°") or rotated in the membrane to be perpendicular to the lipid chains (I called "90°"). I want to see if the protein converge to an orientation, because there are some report that claims that it could stable in both orientations, transmembranal and embeded perpendicular in the membrane.
My protein is coupled with the DPPC residues (tc-grps = Protein DPPC W_Ion). I generate the binary file for runs using an index file, where Protein and DPPC/ Water and Ion are coupled respectively.
I will run again with your sugestions.
I have another question. It is possible to treat just a part of the protein as EN, instead all the molecule? I would like to model the protein with EN, except one C-terminal helix.
Best wishes
Yasser
I am using insane.py script. Basically, I am following the Martini protein tutorials with my membrane protein. My focus is to see whether my protein is inserted in the membrane, comparing standard run with elastic network (standard EN and ElNeDyn). I am running my protein starting with coordinates where the protein is tipically inserted along the TM helices (I called "0°") or rotated in the membrane to be perpendicular to the lipid chains (I called "90°"). I want to see if the protein converge to an orientation, because there are some report that claims that it could stable in both orientations, transmembranal and embeded perpendicular in the membrane.
My protein is coupled with the DPPC residues (tc-grps = Protein DPPC W_Ion). I generate the binary file for runs using an index file, where Protein and DPPC/ Water and Ion are coupled respectively.
I will run again with your sugestions.
I have another question. It is possible to treat just a part of the protein as EN, instead all the molecule? I would like to model the protein with EN, except one C-terminal helix.
Best wishes
Yasser
Please Log in or Create an account to join the conversation.
- mnmelo
-
- Offline
- Admin
Less
More
- Posts: 89
8 years 6 months ago #5027
by mnmelo
Replied by mnmelo on topic Membrane protein insertion in bilayer - mdp files
Hi,
The tc-grps you show go against what I recommended. You should have protein and dppc in the same tc group.
As to a partial elastic network modeling, it is indeed possible, but with some caveats:
-There's no scripting way to do it. You must script it yourself or do it by hand;
-The ElNeDyn topology uses slightly different bonds and angles (the backbone bead is placed at the C-alpha position). You'll have to combine it with a regular Martini topology, which means there might be (small) bonded incompatibilities at the point where the two topologies meet.
This said, I'd warn against mixing topologies, mostly because it's probably unneeded. If you want to give the helix more flexibility it's just better to tweak the elastic network force constant. And don't think a helix will be able to unfold if you remove the network. It won't. It'll probably be as stiff, or more (unless you don't model it as a helix, in which case it will immediately unfold -- secondary structure change is not the strong suit of Martini).
Good luck!
Manel
The tc-grps you show go against what I recommended. You should have protein and dppc in the same tc group.
As to a partial elastic network modeling, it is indeed possible, but with some caveats:
-There's no scripting way to do it. You must script it yourself or do it by hand;
-The ElNeDyn topology uses slightly different bonds and angles (the backbone bead is placed at the C-alpha position). You'll have to combine it with a regular Martini topology, which means there might be (small) bonded incompatibilities at the point where the two topologies meet.
This said, I'd warn against mixing topologies, mostly because it's probably unneeded. If you want to give the helix more flexibility it's just better to tweak the elastic network force constant. And don't think a helix will be able to unfold if you remove the network. It won't. It'll probably be as stiff, or more (unless you don't model it as a helix, in which case it will immediately unfold -- secondary structure change is not the strong suit of Martini).
Good luck!
Manel
Please Log in or Create an account to join the conversation.
- siti
-
- Offline
- Fresh Boarder
Less
More
- Posts: 3
8 years 1 month ago #5397
by siti
Replied by siti on topic Membrane protein insertion in bilayer - mdp files
Hi Manel
You suggested during the equilibration, Berendsen should be used. Does this only for simulation in lipid bilayer? The tutorial for the soluble protein provides the equilibration.mdp using Parrinello-Rahman. Could you please help to explain?
Thanks,
Siti
You suggested during the equilibration, Berendsen should be used. Does this only for simulation in lipid bilayer? The tutorial for the soluble protein provides the equilibration.mdp using Parrinello-Rahman. Could you please help to explain?
Thanks,
Siti
Please Log in or Create an account to join the conversation.
- peterkroon
-
- Offline
- Gold Boarder
Less
More
- Posts: 210
8 years 1 month ago #5407
by peterkroon
Replied by peterkroon on topic Membrane protein insertion in bilayer - mdp files
In general, use berendsen for equilibration, since you are not interested in the distribution during equilibration, and PR is far more likely to blow up.
Please Log in or Create an account to join the conversation.
Time to create page: 0.098 seconds