I am trying to bulid a dense aqueous Lysozyme sysytem by GROMACS. First, I creat a cubic box with side length of 8.7 nm. Then, I put several lysozyme molecules into the box with command 'insert-molecules'. However, due to the van der Waals radii of pair atoms, the maximum number of molecules I can add is 9 which gives a mass concentration of about 300 mg/mL.
Is there any way to put more lysozyme molecules into the box and increase the concentration further?
And, it seems that the literature on the MD of dense protein solutions is rare.Will it cause any problem in the dense case?
Thanks in advance for your advice.
Thanks very much for you suggestion! Qinglu Zhong
Sorry for the late response.Indeed, I have already tried to put 12 lysozyme molecules into the box , but in vain. Although Packmol gave an optimized config, it also reminded that:
Packmol was not able to find a solution to your packing problem with the desired distance tolerance.
IMPORTANT:
If the number of molecules and the restraints are correct, it is still very likely that the current point fits your needs if your purpose is to run a MD simulation. Therefore, we recommend to minimize the energy of the solution found, equilibrate it and run with it as well.
So, I am wondering whether it is rational to start MD with the current configuration obtained from Packmol. If not, is there other way to realize such a box for MD?
You may try : Take the necessary number of proteins in a box of arbitrary size. Then run an NPT simulation of the protein-only system. See if it condenses. You can also increase pressure to get your desired density. Once you get a system of high density, then add water to get your desired density of the solution.
Let me know if this works
没看懂。先用packmol在盒子里随机放12个蛋白，作为初始构象，再在Gromacs里加溶剂。如果这样都出现蛋白放不开或溶剂放不开的情况，就说明盒子尺寸不合适了。
您好! Qinglu Zhong
我是准备把12个溶菌酶蛋白放在一个8.7 nm宽的立方体盒子，我用Packmol试了。一下是我的输入文件：
tolerance 2.0
add_box_sides 1.2
nloop 250
output 1AKI_clean12.pdb
structure 1AKI_clean.pdb
number 12
inside cube 0. 0. 0. 87.
end structure
一般成功运行，packmol会显示类似如下文字是吧：
------------------------------ Success! Final objective function value: .22503E-01 Maximum violation of target distance: 0.000000 Maximum violation of the constraints: .78985E-02 ------------------------------
但是对于我的例子，运行结果显示在此距离限制下没能成功迭代出结果。而且提醒说如果当前给出的结构还是很可能用于MD的，推荐使用目前得到的结果进行能量最小化，弛豫并采样。
现在问题是，我不知道目前这个结果用于MD是否准确，毕竟没有运行得到‘Success!’这样的字段。
不知道您碰到过这种情况没有，谢谢！
Hi Kolattukudy P. Santo Thanks very much for your suggestion!
I have a few questions about your method.Could you please explain it a little bit more?
1.Then run an NPT simulation of the protein-only system. See if it condenses.
------------------------------------->
Did you mean that run the simulation without water?Is that possible?I haven't done that before.And you adviced to run the production MD in NPT ensemble not the NPT equalibration, right?
2. You can also increase pressure to get your desired density.
------------------------------------->
How to deal with that?Proceed with the NPT ensemble?If so, like the former question, shall I release the restraints of protein?
1. Yes, protein-only system with no water. It should be possible, although I havent' simulated no-solvent systems in gromacs, but definitely in AMBER. If you want you can add a little water. The goal is to get a system of your interest.
2. You cannot put restraints on proteins when you want to reach a desired concentration. They will have to move closer.
没有Success是说明确实不好，盒子太小了，或迭代太少了。nloop是否有影响？我没设置过nloop。先试试减少蛋白数量能不能成功？也可以用图形软件看看没成功的结果，只要蛋白间没有重叠就可以继续MD。
Hi Kolattukudy P. Santo
I run the MD following your suggestions and it works now.But I put 5000 water molecules at the beginning, since the unit protein I used has net charge and I have to put some ions to neutralize the system which was accomplished by replacing the water molecules with chloridion.In this regard, I have to put some water molecus ahead.After the EM, NVT and NPT eqilibration(1bar) were carried out, I released the restraints on proteins, run the MD in NPT ensemple(1 bar) and found that the box was compressed to a size slightly lower than the desired value.And it is interesting that the box length always reaches 8.2 nm not matter what pressure is applied.
With the configuration of high density from NPT simulation, I solvated more water molecules and run the MD. It seems nothing abnormal, for instance, no protein is out side of the simulation box and no obvious aggregation was found.However, water density is calculated to be around 1125g/L, which is a little higher than the real physical state.Is it acceptable for simulation?
Thanks a lot for your time!
谢谢您的建议 Qinglu Zhong
nloop 就是迭代次数。Packmol默认是50好像，我增加到了250次，结果仍然不成功。用VMD看了下结构，看不出有什么不好的地方，也没发现明显的重叠。减少蛋白数量的话是可以的，可能packmol也和Gromacs的gmx insert-molecules 命令一样吧，需要充分考虑范德瓦尔斯力。
不过我还是用了这个结构放在Gromacs试了下，当我使用gmx pdb2gmx命令时，出现一下问题：
fatal error: Atom OXT in residue LEU 129 was not found in rtp enty LEU with 19 atoms while sorting atoms.
查了下好像是PDB命名的问题，不知道您碰到没?
Kang Hu : Great that you got things working. Regarding the question on density, you should consider that this is not a pure water system. Moreover, high density protein simulations may lead to artifacts. I suggest that you should carefully analyze this, comparing with simulation/expt data in the literature.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Atomic, Molecular and Optical Physics
- Atoms