MolPy: A Large Language Model-Friendly Toolkit for Reactive Topology Editing in Polymer Simulations
Molecular modeling of synthetic polymers is today increasingly dependent on automation, high-throughput exploration, and the use of a diverse set of simulation tools. However, despite notable progress in molecular simulation methods, the practical infrastructure for constructing, modifying, and parametrizing polymer systems remains fragmented. Here, we launch a toolkit, MolPy, which aims to bridge this gap by providing an extensible and chemically intuitive platform for building polymer systems and orchestrating simulation workflows. The design philosophy emphasizes a clear separation of responsibilities, a stable and inspectable execution model, and a modular architecture that enables flexible extensions. In practice, the framework supports the introduction of new chemical building blocks and bonding schemes that automatically participate in construction, topology generation, and force-field assignment without requiring modification of existing modules. Interoperability with established simulation tools positions MolPy to support streamlined polymer modeling workflows that remain compatible with emerging large language model (LLM)-driven orchestration and other forms of automated pipeline construction.
Authors
- Jichen Li (ORCID: https://orcid.org/0000-0003-3931-6291)
- Wentao Guo (ORCID: https://orcid.org/0000-0001-8058-8323)
- Daniel Brandell (ORCID: https://orcid.org/0000-0002-8019-2801)
- Fabian Schwarz (ORCID: https://orcid.org/0000-0002-6798-3182)
- Ge Sun
Institutions
- California Institute of Technology (US)
- Uppsala University (SE)
- Courant Institute of Mathematical Sciences (US)
- Division of Chemistry (US)
- Tân Tạo University (VN)
- New York University (US)
Publication Details
- Journal
- Journal of Chemical Information and Modeling
- Published
- 2026-06-30
- DOI
- https://doi.org/10.1021/acs.jcim.6c01137
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 0.00