chem-conformer-search

What problem does it solve?

It finds a diverse set of low-energy 3D molecular conformations for a given molecule, so you can prioritize the most physically relevant structures for downstream modeling or calculations.

Core Features & Use Cases

• Conformer generation (RDKit ETKDG): Creates an initial ensemble from a SMILES string or an input structure file.
• MLIP relaxation: Optimizes each conformer with a selectable ML interatomic potential backend (e.g., MACE, MatGL, FairChem) to obtain near-DFT-quality geometries and energies.
• Deduplication and energy ranking with Boltzmann weighting: Removes redundant conformers via RMSD thresholding or clustering (hierarchical/k-means), then ranks and assigns finite-temperature populations.
• Use Case: Given an ibuprofen-like drug with many rotatable bonds, generate a conformer ensemble and extract a compact, low-energy, temperature-relevant set for tasks like docking preparation or conformational preference analysis.

Quick Start

Use the chem-conformer-search skill to generate and MLIP-relax conformers from your SMILES input and write the relaxed ranked ensemble and Boltzmann-weighted results to an output directory.