Schrödinger 2026-3 Release Advances Composite Matrix Design Tools

Source | Schrödinger

Schrödinger (New York, N.Y., U.S.) has released the 2026-3 version of its Materials Science Suite, adding capabilities the company says advance the digital design of composite matrix formulations alongside a new agentic AI tool, machine learning model refinements and expanded structure prediction tools. The release is available for download to Schrödinger users.

A central addition is coarse-grained backmapping, which maps coarse-grained structures to all-atom molecular models. According to Schrödinger, this enables development of structural models for complex resin mixtures, supporting property assessment across length scales relevant to composite matrix formulation.

Featured Content

Bringing Proven FMC Technology to U.S. Manufacturing READ MORE

Episode 55: Dave Noblin, Greene Tweed READ MORE

Automated Composites Knowledge Center READ MORE

The release also updates MS Reactivity workflows, aimed at exploring degradation behavior in polymer systems. The Nanoreactor tool within that workflow now includes a simplified interface with dedicated tabs for workflow and sampling settings, along with a scaled setup for biasing potential, changes that provide more robust tools for assessing thermal and other degradation risks.

Along with the release, Schrödinger also officially introduced Bunsen, a chemistry-native agentic co-scientist now available to early access customers. The tool is designed to execute complex workflows, submit and monitor jobs, and self-correct in real time.

Additional enhancements in the release include a custom extension of Schrödinger’s foundational machine learning force field, which allows users to refine MPNICE force field parameters using custom structure-energy data for specific chemical spaces, and enhanced featurization for solid-state machine learning, which builds machine learning models based on MPNICE embeddings. The release additionally includes extended crystal structure prediction capability, capable of predicting crystal structures for salts, hydrates, solvates and cocrystals.

Visit the company website to learn more about the 2026-3 advances.