Theory and modelling

Shape and stability of self-assembled lipid structures and micro emulsions

We are using the Molecular Dynamics (MD) simulations to model the self-assembly of lipid structures and micro emulsions, and their dependencies on the molecular structure and composition of lipids, temperature, pH, and concentration of cosolvents. The simulations are also used to determine the preferential attachment of the amphiphilic XNA and Ru-tris bpy, as well as the precursors ("food" molecules) of the various chemical components. Atomistic simulations elucidate the spatial arrangement of molecules, which is essential for the correct functioning of the assembly. Experimental work is performed in parallel. As such, the MD simulations are important tool to understand the influence of different chemical derivatives on the spatial structure of the aggregate. In a proceeding project, MD simulations have been performed on the selfassembly and dynamics of a micellar aggregate and on the preferential attachment of a PNA hexamer decorated with hydrophobic anchors to a lipid membrane.

Contact: Harold Fellermann

Calibration of mesoscopic protocell models

Short-time molecular dynamics simulations are used to derive accurate interaction potentials for coarsegrained BD and DPD models by following a procedure developed by Lyubartsev et al. The procedure employs Monte Carlo simulations to obtain effective potentials that reproduce the radial distribution functions of original fully atomistic models. The coarse-grained representation allows for following the system over significantly longer time scales. The results of this calibration are used to develop new model representations of fatty acid, nucleic acid, and sensitizer molecules, as well as to validate and improve existing models of our group that have not been calibrated in full for the time being.

Contact: Harold Fellermann

Scalable mesoscopic simulation techniques with enhanced physical accuracy

To computationally cover the whole mesoscopic range of interest - in particular with respect to modeling vesicles - it is desirable that coarse-grained simulation methods are scalable without introducing serious artifacts. Increasing the coarse-graining of mesoscopic simulations allows for less expensive computations, as it lowers the spatiotemporal resolution of the models. For DPD, an upper coarse-graining limit has been reported that essentially limits the method to the atomistic scale. Introducing alternative scaling relations for the DPD interaction parameters, our group has been able to overcome this coarse-graining limit for the study of static equilibrium properties (e.g. compressibility e.g. miscibility). Further refining the scaling of the DPD thermostat parameters will eventually allow for the correct treatment of dynamical properties (e.g. diffusion). In order to verify the validity of this scaling, large-scale simulations of systems with different coarse-graining parameters have to be performed.

Contact: Harold Fellermann

Replication mechanisms of reverse micelles

Preceding computational studies have proposed an autonomous replication mechanism for micro-emulsions (surfactant-coated oil droplets) as containers for protocells. Experimental evidence suggests that reverse micelles (surfactant-water aggregates in hydrophobic solvent) could outperform the former system. In order to bridge the gap between theory and experiment, a computational model of reverse micelles will be developed and the findings will be compared both with the existing computational model as well as parallel experimental research.

Contact: Harold Fellermann

Non-enzymatic replication of biopolymers

The protocell design of our group relies on non-enzymatic replication of XNA templates by hybridization and ligation of complementary oligomers. It is known experimentally that template replication is thwarted by product inhibition, which ultimately enforces protocellular growth laws that prevent Darwinian evolution. Although preceding theoretical work of our group could show that the implications of product inhibition are globally overcome in our design, unpublished research indicates that product inhibtion nevertheless hinders successful template replication locally, when both template and oligomers are in approximately equal concentration. The project computationally explores ways to overcome product inhibition locally in a simplified DPD model for XNA dynamics.

Contact: Harold Fellermann

Integrated simulations of protocellular life-cycles

Mesoscopic computer simulations, in particular dissipative particle dynamics (DPD), are extended by a stochastic process to model chemical reactions to elucidate the interplay of the arrangement of chemical compounds and reactions. DPD allows us to perform systemic spatially resolved simulations of all key processes in the growth and replication cycle of the aggregate over the relevant time scale. It can therefore elucidate critical aspects of sub-process coupling, ruling out non-working design alternatives and testing hypotheses of more abstract mathematical models. The software has already been developed and we have produced separate results on the container and the genome replication. In the ongoing project, we plan to further combine all key steps in the life cycle of the aggregate within one integrated simulation framework, as well as to refine the model through the ongoing findings both form the experiments and from the other computational subprojects, including reaction kinetics studies.

Contact: Harold Fellermann