Martini Mapper: An Automated Fragment-Based Framework for Developing Coarse-Grained Models within the Martini 3 Framework

TL;DR

Martini Mapper is an automated tool that efficiently generates accurate Martini 3 coarse-grained models from molecular structures, enabling high-throughput simulations of diverse molecules.

Contribution

It introduces a novel, automated framework that constructs Martini 3 models directly from SMILES strings, improving transferability and scalability for complex molecules.

Findings

Generated Martini 3 models for over 6,000 molecules across diverse datasets.

Validated models with transfer free energy benchmarks showing good agreement.

Mapped molecules containing up to 172 heavy atoms, surpassing existing methods.

Abstract

Coarse-graining (CG) reduces molecular details to extend the time and length scales of molecular dynamics simulations to microseconds and micrometers. However, the CG approaches have long been limited by the difficulty of constructing both accurate and transferable models efficiently, considering the large diversity of chemical structures of materials. Among CG force fields, Martini is the most widely used, as it retains essential chemical features while offering substantial computational efficiency. Its most recent version, Martini 3, expands chemical resolution through a much broader bead set, particularly for small molecules. However, this flexibility also complicates the mapping of organic molecules because of context-dependent rules and the lack of standardized procedures. To address this issue, we present an automated framework that builds Martini 3 models directly from SMILES (Simplified Molecular Input Line Entry System) strings by combining a curated bead dictionary with a hierarchical, rule-based algorithm and molecule-specific bonded parameters. Our framework, Martini Mapper https://github.com/eliobaby/Martini_mapper, generated Martini 3 models for 6,280 molecules across six chemically diverse datasets, including 1,689 systems with bond/angle parameters and additional large systems mapped at the topological level. A curated subset of 1,075 mapped structures was benchmarked using transfer free energies in hydrated octanol, hexadecane, and chloroform from water against reference data wherever available. We further examined the benchmark with structural validation via SASA, yielding good agreement with experimental and atomistic reference data. The workflow can also map large molecules containing up to 172 heavy atoms, exceeding the capabilities of existing automated approaches. Our framework, therefore, enables Martini 3 structures for high-throughput simulations.