Megamolecule Self-Assembly\nNetworks: A Combined Computational\nand Experimental Design Strategy

This\nwork describes the use of computational strategies\nto design\nmegamolecule building blocks for the self-assembly of lattice networks.\nThe megamolecules are prepared by attaching four Cutinase-SnapTag\nfusion proteins (CS fusions) to a four-armed linker, followed by functionalizing\neach fusion with a terpyridine linker. This functionality is designed\nto participate in a metal-mediated self-assembly process to give networks.\nThis article describes a simulation-guided strategy for the design\nof megamolecules to optimize the peptide linker in the fusion protein\nto give conformations that are best suited for self-assembly and therefore\nstreamlines the typically time-consuming and labor-intensive experimental\nprocess. We designed 11 candidate megamolecules and identified the\nmost promising linker, (EAAAK)<sub>2</sub>, along with the optimal\nexperimental conditions through a combination of all-atom molecular\ndynamics, enhanced sampling, and larger-scale coarse-grained molecular\ndynamics simulations. Our simulation findings were validated and found\nto be consistent with the experimental results. Significantly, this\nstudy offers valuable insight into the self-assembly of megamolecule\nnetworks and provides a novel and general strategy for large biomolecular\nmaterial designs by using systematic bottom-up coarse-grained simulations.