(33e) Extending the Martini 3 Coarse-Grained Force Field to Polypeptoids

Polypeptoids, also known as poly-N-substituted glycines, are synthetic, sequence-defined polymers and structural isomers of peptides. Unlike conventional peptides, their side chains are attached to the backbone amide nitrogen instead of the α-carbon, eliminating the Cα chiral center and backbone hydrogen bonding. This unique architecture permits both cis and trans configurations of the amide ω-dihedral, leading to increased conformational flexibility and resistance to proteolysis. Depending on side-chain chemistry and solvent conditions, polypeptoids can adopt well-defined three-dimensional structures and further self-assemble into higher-order architectures such as spheres, helices, and sheets. These features have spurred extensive research into their potential roles in drug design, antimicrobial agents, surfactants, and nanomaterials.

Molecular simulations are essential tools for elucidating the behavior of peptoids. While ab initio and all-atom simulations offer high accuracy, their computational cost makes them impractical for large or long-timescale systems. Coarse-grained (CG) models, by reducing atomic detail while retaining key chemical features, significantly enhance simulation efficiency. Among CG frameworks, the MARTINI force field is widely adopted due to its modular design and favorable balance between accuracy and performance. Its latest version, MARTINI 3, offers improved resolution, broader chemical coverage, and enhanced transferability.

In this work, we present a MARTINI 3 compatible CG model for peptoids with various side chains based on a bottom-up parameterization strategy. Atomistic simulations with the CHARMM36 force field were used as references for parameter fitting. To adequately sample the slow cis/trans isomerization of backbone dihedrals, we employed parallel bias metadynamics (PBMetaD). Bonded parameters were derived from atomistic distribution functions via direct Boltzmann inversion (DBI), while nonbonded interactions were primarily adopted from the standard MARTINI 3 parameter library.

The resulting CG model reproduces structural and thermodynamic properties in close agreement with all-atom simulations, while providing significantly enhanced computational efficiency. To facilitate its adoption by the research community, we have integrated the model to MARTINI-based martinize2/vermouth tool, enabling the generation of CG structures and topologies from all-atom structures. This work will establish a robust framework for simulating large-scale peptoid self-assembly, membrane interactions, and nanostructure formation, and supports the rational design of next-generation functional peptoid-based materials.