MAPLE Publications
Papers describing the MAPLE framework and its applications.
Leveraging Machine Learning Interatomic Potentials into Fitting Fast and Accurate AMBER-Compatible Precise Force Field
Introduces ParmFit, an automated MAPLE framework that uses machine-learning interatomic potentials to construct and refine AMBER/GAFF-compatible parameters for small molecules, modified biomolecular residues, and metal coordination environments.
Read on ChemRxiv →
MAPLE: A Machine-Learning Force-Field-Native Platform for Automated Reaction Modeling and Enzyme Design
Presents MAPLE as an MLFF-native molecular modeling platform with a tailored software framework and parallelized algorithms for large-scale, versatile reaction modeling and enzyme design.
Read in Chemical Science →
News & Updates
Release announcements and project updates. See all releases on GitHub.
MAPLE v0.1.5
Added ParmFit workflows for Refinement, NCAA, and MetalAA parameterization, native PDB handling, Direct MaxFlux transition-state searches, FeNNol/FeNNix-Bio1 support, and explicit-solvent compatibility updates.
Release notes on GitHub →
MAPLE v0.1.4
Standardized optimization workflows, corrected optimization metrics and force handling, expanded SCAN methods, improved UMA and MD restart behavior, and introduced calculator plugins and explicit-solvent cluster generation.
Release notes on GitHub →
MAPLE v0.1.3
Expanded MD workflows with NVT/NPT support, GROMACS-style MDP configuration, restart/trajectory improvements, task-aware UMA sizing, MACEPol integration, and IRC/debug cleanup.
Release notes on GitHub →
MAPLE v0.1.1
Added AutoNEB with adaptive image insertion, NVE molecular dynamics, SDCG optimizer with GDIIS acceleration, new IRC integrators (EulerPC, HPC, LQA), rigid scan mode, and input/parser improvements.
Release notes on GitHub →
MAPLE v0.1.0
First public release with frequency analysis, broad ML potential support (ANI, AIMNet2, MACE-OFF23, UMA), geometry optimization, transition state search, IRC, PES scan, solvation, constraints, GPU acceleration, and D4 dispersion.
Release notes on GitHub →