Comparative Molecular Dynamics Characterization of Hair Keratin Unfolding Mechanics

Keratin proteins are fundamental structural components of hair fibers, contributing to their mechanical resilience, elasticity, and fracture resistance. However, systematic molecular-scale characterization of keratin unfolding mechanics across protein types remains limited, restricting the connection between protein-level deformation mechanisms and hierarchical hair fiber mechanics. Here, we establish a comparative molecular-dynamics-based framework for characterizing the unfolding behavior and nanomechanical response of a curated dataset of 51 keratin proteins. We conduct implicit atomistic molecular dynamics (MD) simulations, including equilibration and steered molecular dynamics (SMD) under four accelerated pulling velocities, to quantify unfolding forces, energy absorption, and structure-property relationships. These accelerated pulling conditions are interpreted as computational probes of relative molecular-scale trends, rather than direct reproductions of experimental hair-fiber strain-rate regimes. Across these accelerated SMD conditions, the simulations show rate-sensitive increases in unfolding force and energy absorption, consistent with constrained molecular relaxation during faster molecular pulling. Stronger correlations between nanomechanical properties and molecular descriptors emerge at higher pulling rates, and the nanomechanical responses of different keratin types (Type I and II) are also compared. The findings provide molecular-level insights into protein unfolding mechanisms that may contribute to the mechanical behavior of hierarchical keratin structures. This study establishes a quantitative framework for comparative keratin unfolding mechanics, providing molecular-level descriptors for future multiscale modeling of hair fiber behavior. These results support applications in biomaterial design, hair fiber durability analysis, and bioinspired material engineering. Future work will integrate these nanomechanical descriptors with fiber-level mechanics and machine learning-based keratin design.