Hello everyone,
I currently have a protein in a fully equilibrated membrane system. For the analysis I am running, I need to remove Na+ ions from the inside of the pore region. I would normally do a distance restraint, but I have ~191 Na+ ions within my box. I was told that adding a distance restraint, on each of the Na+ ions would make the simulation run prohibitively slow.
What would be the best way to keep ions out of the pore within GROMACS? I have considered a cylindrical restraint but it looks like that extends all the way through the entire z axis which would mess up the electric double bilayer that has naturally formed and is important.
Thanks for any insight!
Dear Hannah Burton
What about a flat bottomed potential along z centered in the membrane? Like here:
https://manual.gromacs.org/2024.2/reference-manual/functions/restraints.html#flat-bottomed-position-restraints
You have to make it less large than the membrane otherwise you will push away the ions on the membrane surface, but it should be more that sufficient to keep any ion to go through the channel. Or you could similarly implement a spherical restraint and put it in the center of the channel.
Hi I checked this out and it seems like I could only specify the spherical restraint on each individual ions to keep them in a radius surrounding them. Do you know how I can apply the location of the sphere to the protein to repulse ions away from it?
Hi Nicola Piasentin thanks for the response. I have thought about it a bit more but would really like some guidance as to how I could implement this as I am new to GROMACS and there is no tutorial for this particular restraint-type.
In order to do this, would I just pass grompp a -r flag with non-physical coordinates on the Na+ ions? So I would place every single Na+ ion at, say, the COM of the protein in the reference file and then create a spherical flat bottomed restraint so they never actually enter into the region? Do I need to specify a negative k value to keep them out?
Dear Hannah Burton ,
exactly, what you are doing is giving to GROMACS the restraint file with -r and from there define the regions where you do not want the ions to go. In this thread there is some info about the flat bottom restraint
https://gromacs.bioexcel.eu/t/methodology-for-flat-bottomed-position-restraints/9366
I guess two options for you could be
i) You calculate the center of mass of your membrane along z (I am assuming you have a bilayer along the xy plane) and project the z values of your Na+ ions of your box on that value. Then in the topology of the Na+ ions you simply add a line that may look something like this
#ifdef flat_bottomed_restraint_name
[ position_restraints ]
1 2 5 radius force_constant
#endif
ii) You calculate the center of you channel in (x,y,z) and project the (x,y,z) values of your Na+ ions of your box on that value. In this case in the topology of the Na+ ions you simply add a line that may look something like this
#ifdef flat_bottomed_restraint_name
[ position_restraints ]
1 2 1 radius force_constant
#endif
So in case i) all you reference positions for Na+ ions will belong to the plane passing through the lipid bilayer COM and you will keep them away from there with a planar flat bottomed restraint along z. In this case r will be negative and I would pick a value that is less than the bilayer thickness so you do not mess (too much) with the electrostatics of the bilayer surface. Having a plot of the density of Na+ along z for an unbiased run can give you a good idea of how large this value of r can be. Regarding the k value, I would keep it small, like 10/20 kJ mol-1 and see if that is sufficient.
In case ii) I would do the same, as the radius should probably still be smaller than the bilayer thickness (or better take a look at the distribution of Na+ ions) so you do not mess too much with the electrostatics of the surface.
Just remember in both cases to have a negative r value otherwise you will pull all the ions in that position, which will destroy for sure your box and or kill your simulation!
There are other fancier ways to implement these restraints, e.g., you may put a repulsive potential via a dummy atom in the center of the channel and/or add fancy funnels by using external tools such as colvar (https://colvars.github.io/pdf/colvars-refman-gromacs.pdf) or PLUMED (https://www.plumed.org/doc). Also take a look at other people doing the same in previous literature so that you can have an idea if someone implemented these restraints and if any further consideration should be made.
Good luck with your sims :)
- Badges
- Science topic
- Similar topics
- Filing