Fuzzy cluster modes in micellar dynamics

Micellar dynamics show various fluctuation shapes which may reflect certain physical processes. We classify the fluctuation shapes by combining physical meaning of symmetry and isotropy. The symmetry and isotropy correspond to roughness and parallel degree, respectively. Hence, their couplings indicate various possible shape modes. To generate various micellar clusters, we perform molecular dynamics simulation for small number of lipids of POPC and POPE at various conditions such as 300 K, 340 K, and both pure and salty water. Beside easily distinguishable zones, we find indistinguishable ones. The indistinguishable or fuzzy zones indicate fuzzy cluster modes and a shape transition possibly has responsible for the modes. To reveal the fuzzy zones, we investigate the modes by applying fuzzy c-means clustering. The membership degree properties of the method enable to analyze smooth transition shapes. We find not only smooth classification but also more specific values in the fuzzy cluster modes.