pdb2reaction: End-to-End Reaction-Path Elucidation from PDB Structures Using Machine-Learning Interatomic Potentials

Elucidating enzymatic reaction mechanisms requires a sequence of computational tasks comprising active-site extraction, minimum-energy path (MEP) search, transition-state (TS) refinement, intrinsic reaction coordinate (IRC) validation, and quasi-rigid-rotor harmonic-oscillator (QRRHO) thermochemistry—stages typically connected by ad hoc scripting and per-system tuning. We present pdb2reaction, an open-source Python command-line toolkit that automates this entire pipeline directly from a user-curated PDB using a single machine-learning interatomic potential (MLIP) backend. A GPU-accelerated pysisyphus fork is bundled to perform all Hessian-based heavy computations on the same CUDA device as the MLIP, minimizing data transfer overhead. This toolkit also implements a bond-change-driven recursive path-search algorithm, which successfully recovered the two-step reaction mechanism of the geranyl pyrophosphate (GPP) C6-methyltransferase BezA from a single reactant PDB and a minimal scan list. The identified mechanism comprises an S N 2-like methyl transfer followed by glutamate-mediated deprotonation via a cationic intermediate. A benchmark comprising 23 reaction steps across six different enzymes demonstrated that pdb2reaction recovered the literature mechanisms for 16 out of 23 steps using two of the five tested MLIP backends, with each successfully yielding transition states characterized by a single imaginary frequency. pdb2reaction enables rapid PDB-tomechanism elucidation on a single GPU, providing a high-throughput complement to higher-cost higher-cost DFT-based cluster-model calculations and full enzyme–solvent QM/MM calculations.