Maple
MAchine Learning Potential for Landscape Exploration (MAPLE)
MAPLE is a powerful computational chemistry toolkit that leverages machine learning potentials for efficient structure optimization, transition state searching, and reaction pathway analysis.
Overview
MAPLE integrates state-of-the-art machine learning models (AIMNet2, ANI) with classical optimization algorithms to enable:
- Structure Optimization: LBFGS, RFO algorithms for finding energy minima
- Transition State Search: NEB (Nudged Elastic Band), CI-NEB, String Method, Dimer Method
- Reaction Path Analysis: IRC (Intrinsic Reaction Coordinate) calculations
- Vibrational Analysis: Frequency calculations with mass-weighted Hessian
The software is designed for computational chemists who need fast, accurate quantum-mechanical calculations without the computational cost of traditional ab initio methods.