- Posts: 5
self-assembling peptides
- saracino
-
Topic Author
- Offline
- Fresh Boarder
Less
More
8 years 3 months ago #5359
by saracino
self-assembling peptides was created by saracino
Hi all,
I'm simulating 100 self assembling peptides in water at 1%w/v concentration.
The peptides are formed by 11 residues, the starting conformations aren't extended but derived by a united atom conformational sampling of the monomer.
The peptides are randomly distributed in the box, periodic boundary conditions are used and parameters for extended conformation are assigned.
To avoid problems with the genbox algorithm in adding water from boxes of dimensions smaller than the desired ones, new water boxes have been prepared and equilibrated on purpose.
Following the Peptide-assembly tutorial available on the Martini website I started with the 5000 minimization steps. At the end of the minimization, the systems showed the formation of vacuum bubbles and jagged edges that in turn were completely reabsorbed in the equilibration phase. As the peptides showed a filamentous mechanism of self-assembly we was wondering how it may be inducted/influenced by the vacuum and indentations resorption at the beginning of the equilibration phase.
In the tutorial the equilibration phase is a simulation used also for the production. To allow the disappearance of the bubbles and indentations, I decided to try the introduction of an intermediate equilibration phase to allow the solvent (water and ions) to relax, that is a 500ps of md with position restraints on the peptides before starting the real production phase. Often this phase have difficulties to proceed and end with segmentation faults signals.
Thinking that this may be related with the formation of the vacuum bubbles also the densities of the system have been checked.
The result was that in our 100 peptide system the water filling procedure performed by genbox with rvdwd 0.21 introduces about ten thousand water molecules below the forth part of the W molecules introduced in the case of the united atom version of the same system.
This seems a bag in the filling procedure.
I tried to solve the problem introducing a number of W molecules comparable to the fourth of the ones introduced in the corresponding united atom system, this required the use of a vdwd 0.14 in genbox. After that, less important vaccum zones are formed and the position restrained relaxation and the production go on without difficulties.
Another way may be to run a second time genbox after a first minimization to fill the bubbles but this way isn't easy to control, it's easy to reach a too much high density.
So I have the following questions:
- is it expected the vacuum bubbles formation after the minimization?
- during the equilibration/production, may the bubbles resorption affect the next self-assembly mechanism?
- have you already had problems in using genbox with Martini for highly concentrated systems?
- is it good to run a PR simulation to relax the solvent before starting the production or is this procedure that may introduce some artefact in the self assembly mechanism?
Thank you in advance,
Gloria
I'm simulating 100 self assembling peptides in water at 1%w/v concentration.
The peptides are formed by 11 residues, the starting conformations aren't extended but derived by a united atom conformational sampling of the monomer.
The peptides are randomly distributed in the box, periodic boundary conditions are used and parameters for extended conformation are assigned.
To avoid problems with the genbox algorithm in adding water from boxes of dimensions smaller than the desired ones, new water boxes have been prepared and equilibrated on purpose.
Following the Peptide-assembly tutorial available on the Martini website I started with the 5000 minimization steps. At the end of the minimization, the systems showed the formation of vacuum bubbles and jagged edges that in turn were completely reabsorbed in the equilibration phase. As the peptides showed a filamentous mechanism of self-assembly we was wondering how it may be inducted/influenced by the vacuum and indentations resorption at the beginning of the equilibration phase.
In the tutorial the equilibration phase is a simulation used also for the production. To allow the disappearance of the bubbles and indentations, I decided to try the introduction of an intermediate equilibration phase to allow the solvent (water and ions) to relax, that is a 500ps of md with position restraints on the peptides before starting the real production phase. Often this phase have difficulties to proceed and end with segmentation faults signals.
Thinking that this may be related with the formation of the vacuum bubbles also the densities of the system have been checked.
The result was that in our 100 peptide system the water filling procedure performed by genbox with rvdwd 0.21 introduces about ten thousand water molecules below the forth part of the W molecules introduced in the case of the united atom version of the same system.
This seems a bag in the filling procedure.
I tried to solve the problem introducing a number of W molecules comparable to the fourth of the ones introduced in the corresponding united atom system, this required the use of a vdwd 0.14 in genbox. After that, less important vaccum zones are formed and the position restrained relaxation and the production go on without difficulties.
Another way may be to run a second time genbox after a first minimization to fill the bubbles but this way isn't easy to control, it's easy to reach a too much high density.
So I have the following questions:
- is it expected the vacuum bubbles formation after the minimization?
- during the equilibration/production, may the bubbles resorption affect the next self-assembly mechanism?
- have you already had problems in using genbox with Martini for highly concentrated systems?
- is it good to run a PR simulation to relax the solvent before starting the production or is this procedure that may introduce some artefact in the self assembly mechanism?
Thank you in advance,
Gloria
Please Log in or Create an account to join the conversation.
- Pim
-
- Offline
- Expert Boarder
Less
More
- Posts: 105
7 years 10 months ago - 7 years 5 months ago #5653
by Pim
Replied by Pim on topic self-assembling peptides
Hi Gloria,
This is not a very fast reply, but maybe it will help you still.
First of all: genbox has problems with Martini water, it's been a strange bug that gives incorrect numbers of water molecules that's been annoying me for years, but it seems it has finally been solved in Gromacs 5.1.1 and onwards (I used 5.1.2). Even if you still want to run your simulations with other versions, I'd recommend using this to create your systems.
Second, if you have many small molecules in your system, the procedure of genbox indeed often adds to few water molecules simply because you have many irregular voids (using -vdwd 0.2 is already a bit better than 0.21 in this case). This creates some vacuum bubbles if you minimize for a long time or if you simulate without pressure coupling. The solution: don't minimize too long and use pressure coupling :). For a simulation you describe I sometimes don't even minimize at all! That's not general good practice but it helps sometimes.
I would not use the position restraints on the peptides in your case, it will likely create more problems than without it, or at least use "refcoord_scaling all". The thing is that there are simply not enough water molecules for the box size, that's not going to solve itself unless you allow the box to shrink.
Best of luck! I should also update the peptide tutorial some time soon to state the above.
This is not a very fast reply, but maybe it will help you still.
First of all: genbox has problems with Martini water, it's been a strange bug that gives incorrect numbers of water molecules that's been annoying me for years, but it seems it has finally been solved in Gromacs 5.1.1 and onwards (I used 5.1.2). Even if you still want to run your simulations with other versions, I'd recommend using this to create your systems.
Second, if you have many small molecules in your system, the procedure of genbox indeed often adds to few water molecules simply because you have many irregular voids (using -vdwd 0.2 is already a bit better than 0.21 in this case). This creates some vacuum bubbles if you minimize for a long time or if you simulate without pressure coupling. The solution: don't minimize too long and use pressure coupling :). For a simulation you describe I sometimes don't even minimize at all! That's not general good practice but it helps sometimes.
I would not use the position restraints on the peptides in your case, it will likely create more problems than without it, or at least use "refcoord_scaling all". The thing is that there are simply not enough water molecules for the box size, that's not going to solve itself unless you allow the box to shrink.
Best of luck! I should also update the peptide tutorial some time soon to state the above.
Last edit: 7 years 5 months ago by Pim. Reason: gmx version
Please Log in or Create an account to join the conversation.
- beenish
-
- Visitor
7 years 6 months ago #5956
by beenish
Replied by beenish on topic self-assembling peptides
Hi Gloria
I am working on similar kind of project. I was just wondering, could you see an ordered fibrillar formation? For example, I am working on self assembly of beta amyloid peptides and according to the previous studies it should give a well organized, ordered fibril (which I can not produce, no matter what I do ;( ) I thought it may be because of lack of directional hydrogen bonding.
Thanks
I am working on similar kind of project. I was just wondering, could you see an ordered fibrillar formation? For example, I am working on self assembly of beta amyloid peptides and according to the previous studies it should give a well organized, ordered fibril (which I can not produce, no matter what I do ;( ) I thought it may be because of lack of directional hydrogen bonding.
Thanks
Please Log in or Create an account to join the conversation.
Time to create page: 0.091 seconds