MARTINI model, lipids

Lipid phase transitions are often slow on the timescale accessible to simulations, such as the gel to liquid-crystalline phase transition. Using the MARTINI model, liquid-crystalline to gel phase transitions and domain formation have been... 

It is also important that the potential is not so strong that the system can solidify at physiological conditions. For example, the MARTINI water model is known to spontaneously crystallize at physiologically relevant temperatures [17]. This phenomenon is enhanced by the presence of interfaces (e.g., a lipid bilayer surface), and requires the addition of unphysical anlilieeze particles to avoid crystallization. While we note that modifications to the MARTINI water model exist (e.g., adding charge polarization) [41, 42], only the original MARTINI model was tested, since it more closely resembles the model derived via MS IBI (i.e., both... 

Fig. 1 Illustration of a buckling simulation using the MARTINI model for DMPC (which uses 10 heads pCT lipid)...

In particular, this parameterization approach has shortcomings in reproducing correctly the phase diagram of water, including, for example, water/vapor interfacial tension, and, as a consequence, it has flaws for what pertains the investigation of surfactant properties of lipid monolayer at the airAvater interface.Improvements of the MARTINI water model have been proposed via the use of a polarizable force field, but polarizable CG force fields remain poorly investigated so far in the field of CG simulations. 

Another chemically accurate CG force field for MD simulations of lipid assemblies is the Shinoda-DeVane-Klein (SDK) foree field. This model is the improved version of the earlier CG model developed by Shelley et al. and utilizes a three heavy atoms to one CG bead mapping, thus providing a slightly higher resolution than the MARTINI (Fig. 12). 

As it is expected from its finer mapping and more sophisticated parameterization procedure, this model is generally more accurate than MARTINI, especially for what pertains properties at the lipid-water interface. On the other hand, the unusual functional form of Lennard-Jones interactions (that is not supported by most current MD software) and the need for a careful parameterization every time a new molecule is introduced, make this force field less appealing to the non-initiated. 

Nevertheless, the authors have reported so far a large range of applications on lipid and membrane systems of sizes comparable to those studied using MARTINI, including spontaneous vesicle self-assembly, monolayer collapse, vesicle to micelle transformation and membrane fusion. The SDK model thus represents a powerful and promising approach to investigate finely tuned chemical properties of large lipid assemblies when cheaper but less accurate CG models turn out to be unsuccessful. 

Because of system sizes, these investigations are clearly not feasible with atomistic MD simulations for example a lipid vesicle of a radius of approximately 25 nm would contain around 18 000 lipids, corresponding to approximately 2 500 000 atoms using a fully atomistic force field, even before taking into account the unavoidable presence of water molecules. However, thanks to progresses both in hardware and in software technologies, such systems are nowadays computationally tractable using CG models such as the MARTINI or the SDK. 

---