Abstract:

Active molecular self-assembly is a fundamental process that drives the formation of complex structures from simple building blocks in response different environmental factors. This dissertation employs computational approaches to investigate the molecular mechanisms governing active self-assembly in bio-inspired molecules, and computationally design new candidates. Multi-scale molecular dynamics simulations of the CSH-CSSC hydrogel system reveal that both oxidized (CSSC) and reduced (CSH) forms participate in aggregation, with CSSC driving assembly while CSH contributes to stability. Deep learning accelerated high throughput coarse grained MD simulations, coupled with assembly network statistics analysis, enabled screening of large sequence space of cysteine-containing peptides, and produced models ready for future refinements.