List view
Overhauls the internal parameter generation engine to standardize all force field terms, ensuring rigorous alignment with the `dreid-kernel` physics model. The release corrects critical parameterization logic for dihedral and inversion terms and introduces support for energy reflection in the Exp-6 van der Waals potential, implementing Newton's method for calculating local maximums to guarantee smooth short-range repulsion.
Due by February 21, 2026•6/6 issues closedElevates dreid-forge from a general-purpose parameterizer into a specialized engine for bio-organic hybrid systems. This release fundamentally redesigns the electrostatic treatment, enabling distinct charge strategies for biological macromolecules, small molecule ligands, and ions within a single system. It integrates the `ffcharge` library to access standard AMBER and CHARMM charge sets and leverages `cheq` v0.5’s new external field capabilities to compute polarized ligand charges in response to a protein environment.
Due by January 6, 2026•1/1 issues closedTransforms dreid-forge from a foundational library into a robust, compositional command-line utility. This version introduces a unified CLI that makes the automated DREIDING parameterization engine accessible directly from the terminal. It provides seamless I/O for diverse chemical formats (PDB, SDF, MOL2) and fully supports Unix-style piping, enabling effortless integration into complex molecular dynamics workflows. The CLI exposes granular controls for atom typing, QEq charge equilibration, and potential function generation, bridging the gap between static molecular structures and dynamic simulation inputs.
Due by December 13, 2025•1/1 issues closedEstablishes the foundational release of dreid-forge, delivering a high-performance orchestration engine for automated DREIDING force field parameterization across both biological and chemical domains. This version introduces a unified, type-safe system representation that seamlessly bridges diverse molecular inputs with rigorous physics-based simulation parameters. The core pipeline integrates a context-aware topology perception module for precise hybridization and resonance assignment, coupled with a self-consistent Charge Equilibration (QEq) solver for dynamic electrostatic modeling. Furthermore, the release features an adaptive parameterization engine that synthesizes explicit valence terms and torsional barriers directly from first-principles hybridization rules, enabling the rapid generation of production-ready LAMMPS inputs for proteins, nucleic acids, and arbitrary small molecules.
Due by December 8, 2025•4/4 issues closed