banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Biomolecular simulations of membranes: Physical properties from different force fields
Rent this article for


Image of FIG. 1.
FIG. 1.

Schematic drawing of the DOPC lipid and the naming convention used in this manuscript.

Image of FIG. 2.
FIG. 2.

Area per lipid for the five investigated systems in simulation time smoothed by a sliding window of length. Larger fluctuations were observed for systems applying surface tension (B, C, and D) as compared to those using semi-isotropic pressure coupling (A and E). Note that the experimental area per lipid for DOPC was measured to be at (Ref. 79).

Image of FIG. 3.
FIG. 3.

Deuterium order parameters, , for the two acyl chains of DOPC lipids analyzed for systems B, D, and E. Error bars indicate the errors of the means from block averaging (blocks of length). Experimental values are given for the POPC (black squares) and chain (gray squares) at (Ref. 51), and for DOPC at (Ref. 53) (◇ symbols). Note that the double-bonded carbon atoms are at positions 9 and 10.

Image of FIG. 4.
FIG. 4.

The (symmetrized) electron density profiles of the overall and individual chemical components of all simulation systems. The profiles are centered at the core of the bilayer, and the standard errors (as shown here in gray lines) are calculated by dividing the trajectories into blocks of length. The profile was computed by placing the appropriate number of electrons at the center of atomic nuclei binning along the direction of the membrane normal (bin width of ). The experimental density profile (Ref. 79) is shown as a dashed line.

Image of FIG. 5.
FIG. 5.

Comparison of the electron density profiles at the membrane-water interface (centered on the lipid bilayer) for force field combinations GAFF-SPC/E, CHARMM27-TIP3P, and Berger-SPC (systems B, D, and E). The same coloring scheme for Fig. 4 is used here.

Image of FIG. 6.
FIG. 6.

Lipid diffusion coefficients calculated for different time lengths of the simulations by block averaging: The trajectory was divided into pieces and the msd (upper panel) was calculated separately for each block. Diffusion coefficients (lower panel) computed for different time ranges were obtained by fitting different time windows of the msd curve. For the short-range diffusion (colored ◇ symbols), shorter time windows of (fit starts between 10 and ) and (fit starts between and ) were used. For the long-range diffusion (colored ☉ symbols) fitting was done on the linear segment of the msd curve between 5 and .

Image of FIG. 7.
FIG. 7.

The lateral diffusion coefficients of water molecules (solid lines) at the membrane-water interface. The dashed lines mark the normalized DOPC electron density profiles (superimposed at the descending slope) to indicate the location of the bilayer. The respective values for bulk water diffusion are shown as dashed-dotted lines (errors in gray shading). All diffusion coefficients (and standard errors) were measured by fitting the slope between 2 and of the msd curves, using block averaging of length .

Image of FIG. 8.
FIG. 8.

Total electrostatic potential (upper panel) and contributions (lower panel) due to the lipid dipoles (solid lines) and due to water dipole orientation (dashed lines) at the membrane/water interface across the DOPC bilayer (symmetrized). The potential was averaged over the final of the simulation.

Image of FIG. 9.
FIG. 9.

Average water dipole moment along the membrane normal. The large fluctuations inside the hydrophobic core of the lipid bilayer are due to individual water molecules spontaneously permeating the membrane.


Generic image for table
Table I.

The atom types and the partial charges of all atoms used in the respective force fields.

Generic image for table
Table II.

Overview of the simulation systems presented in this study.

Generic image for table
Table III.

Averages and errors of the membrane structural parameters calculated by block averaging (block length of ). The membrane thickness was measured as the distance between the two peaks in the system electron density profiles. Errors in the area per lipid for systems D and E were rounded to .

Generic image for table
Table IV.

Distribution of lipid headgroup orientations computed as the angle between the vector connecting the phosphorus and the nitrogen atoms and the bilayer normal. Both the maximum of the distribution—the most probable orientation—and its width are given. The last two columns give the most probable total dipole moment of the individual lipids in the respective force fields (± the width of a fitted Gaussian distribution) as well as the average of the dipole moment along the bilayer normal (-direction).

Generic image for table
Table V.

The calculated mean lateral diffusion coefficient for water molecules and for DOPC molecules. For water diffusion, the average was calculated from all water molecules in the system regardless of the distance to bilayer. For lipid diffusion, the long-range diffusion coefficients are shown here while the short-range diffusions are depicted in Fig. 6.


Article metrics loading...


Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Biomolecular simulations of membranes: Physical properties from different force fields