I’m trying to get a mixed lipid bilayer using the Slipids force field to pack closer together. The method I used to insert my lipids was manual, and I tried to arrange them as neatly as possible without any neighbouring lipids getting too close. The bilayer looks good, and I’ve successfully minimised it.
However, when I solvate and start running NVT, my lipid tails collapse, creating a vacuum in the middle of the bilayer. I believe this happens because my lipid molecules aren’t packed tightly enough and there are some vacuum gaps.
I can force them to stay in place with heavy position restraints, and the bilayer becomes more compact, but then the water on either side remains in the same xy-dimensions, which gives the system an hourglass shape and looks wrong!
Have you tried CHARMM-GUI? It should support also the slipids force field, and will make your life extremely easier in terms of setting up a lipid bilayer.
I have some lipids which are only newly parameterised in literature so they have been specially generated with special .itp files for the SL lipids forcefield (such as some larger surfactant-like molecules) and I can’t source them via CHARMGUI unfortunatley!
I can recommend PACKMOL-Memgen from AmberTools (here). You can easily add novel structures to the database to build lipid bilayers. Depending on the exact structure of your lipid, you probably have to fiddle around with the tool’s parameters. For the equilibration of the generated membrane you can either use your own workflow or copy the one from the CHARMM-GUI.