Multiscale computational prediction of β-sheet peptide self-assembly morphology

Although nanostructures self-assembled by short peptides are very promising in developing novel biomaterials and nanomaterials, it is still a great challenge to design the peptide molecular structure which will self-assemble into designated nanostructures. By combining elastic theory with molecular dynamics simulations, we introduce a multiscale computational approach to predict the beta-sheet morphology self-assembled by short peptides in aqueous solution only based on the molecular structure of the peptide. In our approach, the gap between the elastic model and atomistic model is bridged by the simplified model, whose parameters are determined by enhanced sampling and extensive all-atom molecular dynamics simulation results at different levels. This multiscale approach is applied to two model peptides KIIIIK (KI4K) and IIIIKK (I4K2) to test its validity. The computational results, consistent with the previous experimental observations, show that KI4K with a higher ratio of inter-sheet interaction to intra-sheet interaction tends to form tube-like morphology with a larger width, while I4K2 with a lower ratio tends to form fibril with a smaller width. This methodology is anticipated to be helpful for computer-aided design of nanostructures self-assembled by short peptides.

Automated summary

A multi-scale computational approach has been introduced to predict the beta-sheet morphology self-assembled by short peptides in aqueous solution based on the molecular structure of the peptide. The computational results confirm that peptides with higher inter-sheet interaction tend to form tube-like structures with larger width, while peptides with lower ratios tend to form fibril structures with smaller width, which is consistent with experimental observations.