# Description: # This constructs a bilayer constructed from coarse-grained DPPC lipids # (implicit solvent). The DPPC lipid model is described here: # G. Brannigan, P.F. Philips, and F.L.H. Brown, # Physical Review E, Vol 72, 011915 (2005) # # NOTE: There is an example of a 50%/50% DPPC & DLPC mixture # in the "membrane+protein" and "vesicle" examples. import "CGLipidBr2005Orig.lt" using namespace CGLipidBr2005 lipids = new DPPC [32].move(7.5, 0, 0) [37].move(3.75, 6.49519, 0) [2].rot(180, 1, 0, 0) # Move the lipds up to the center of the box lipids[*][*][*].move(0,0,75.0) # Although this patch of lipids is not square or rectangular, (it looks # like a parallelogram), this is no longer the case after rectangular # periodic boundary conditions are applied. We apply them below: # width: 240 = 32*7.5 # height: 240.322 = 37*6.49519 write_once("Data Boundary") { 0 240 xlo xhi 0 240.322 ylo yhi 0 150.0 zlo zhi } # -------------- File ends here. Only comments below.------------------- # ------------------------------------ # ------------- COMMENTS: ------------ # ------------------------------------ # # A note on geometry: # We want to create a bilayer arranged in a hexagonal lattice consisting of # 32 rows (each row is aligned with the x-axis) # 37 columns (aligned at a 60 degree angle from the x axis) # The lattice spacing is 8.0 Angstroms ( ~= 0.95*sigma*2^(1/6), where sigma=7.5) # When wrapped onto a rectangular box, the dimensions of the system are: # 32 * 7.5 Angstroms in the X direction # 37 * 7.5*sqrt(3)/2 Angstroms in the Y direction # ------------------------------------ # # Below I show simple ways to create a lipid bilayer: # # 1) If you just want to make lipid bilayer out of DPPC, # without specifying the location of each lipid, you could use this syntax: # lipids = new DPPC [32][37][2] # 3-D array # Later you can load in the coordinates of the lipds from a PDB file. # Alternately you could also use a 1-dimensional array: # lipids = new DPPC [2368] # 1-D array. Note: 2368 = 32 x 37 x 2 # It does not matter as long as the number of lipids is correct. # Multidimensional arrays are only useful if you plan to apply independent # coordinate transformations to each row and column and monolayer. See below: # # 2) Instead of loading a PDB file later, we can directly specify the location # of each DPPC lipid in the LT file itself. For lipid bilayers, this is # easy, because the bilayer structure resembles 2 planar lattices. # We can use "move" commands to place each lipid, and the "rot" command # to turn the lipids in one of the monolayers upside down. # # lipids = new DPPC [32].move(7.5, 0, 0) # [37].move(3.75, 6.49519, 0) # [2].rot(180, 1, 0, 0) # # 3) If you want to create a bilayer from a mixture of DPPC and DLPC, you must # replace "DPPC" in the command above with random([DPPC,DLPC],[0.5,0.5],12345) # Here "0.5,0.5" are the probabilities for each molecule type, and "12345" # is an optional random seed. # lipids = new random([DPPC,DLPC], [0.5,0.5], 12345) # [32].move(7.5, 0, 0) # [37].move(3.75, 6.49519, 0) # [2].rot(180, 1, 0, 0) #