bioinfkaustin
commented
5 months ago Hi @LiangYat-sen, thank you very much for your comment! I'm not an expert in peptide self-assembly, but I'll answer as best as I can.
If you are able to put together a GROMACS input system that addresses your question about self-assembling peptides, then I think our Production Notebook would be appropriate for running that system on a cloud GPU service. (You would upload that GROMACS system to your Google Drive, and then have the Production Notebook use that folder as its working directory. The notebook will ensure that the simulation runs on the GPU automatically if one is present, as long as you don't choose any simulation parameters which are incompatible with the GROMACS GPU code -- which would be rare in recent versions of GROMACS.)
It's possible that CHARMM-GUI could be useful for constructing such a system, and that therefore, the CHARMM-GUI Notebook could also help you. One suggestion is that you might use their Multicomponent Assembler tool to make a system containing the desired number of peptides -- which would hopefully come together and assemble into a bulk material during the simulation.
One piece of advice which may be useful to know -- changes in protonation state are often important in peptide folding and conformational change. One approach to account for this is to manually adjust the protonation state of your amino acid chain at the beginning of the simulation. However, that is only a partial solution, because these states will not dynamically vary as they would in reality. For example, in a standard GROMACS simulation the protonation state of the side-chain of histidine residues is fixed to match the input file, but in the real world, histidine residues fluctuate between these states (at physiological pH). This is a limitation in GROMACS, but it has been addressed with a fork for so-called "constant pH molecular dynamics". It should be straightforward to modify these notebooks to use an installation of this fork rather than of standard GROMACS if this is important for your research question.
Finally, I should say that the simulation of covalent bonding is probably not possible with this system. If your self-assembling peptide's mechanism of assembly relies on forming covalent bonds, you would need to incorporate quantum mechanics into your simulation to see that in the simulated trajectory. GROMACS can integrate with CP2K to undertake a mixed QM/MM simulation, but this is careful work and these simulations tend to be very demanding -- I would suggest collaborating with someone who has expertise in this kind of system if you need to address that kind of question.
Hope this helps! :-)
Hello, I would like to ask if this file can be used for peptide self-assembly.What do I have to do if I want to use the GPU version. Thank you for your answer