HTOCSP

Official Resources

• Homepage: https://github.com/MaterSim/HTOCSP
• Source Repository: https://github.com/MaterSim/HTOCSP
• Documentation: https://github.com/MaterSim/HTOCSP/blob/main/README.md
• License: MIT License

Overview

HTOCSP is a Python-based framework specifically designed for the automated, high-throughput prediction of organic crystal structures. It integrates various open-source tools to generate, optimize, and rank crystal structures of organic molecules, addressing the challenge of polymorphism in pharmaceutical and organic materials.

Scientific domain: Organic crystal structure prediction, high-throughput screening
Target user community: Pharmaceutical researchers, organic chemists, computational materials scientists

Theoretical Methods

• Random Structure Generation: Uses PyXtal for symmetry-compliant generation.
• Force Field Optimization: Uses CHARMM/Amber force fields via LAMMPS or GULP for initial screening.
• DFT Optimization: Refinement using DFT (e.g., VASP, Quantum ESPRESSO).
• Clustering/Ranking: Structure matching to identify unique polymorphs.

Capabilities

• Automated Workflow: From SMILES/molecule to ranked crystal structures.
• Symmetry Handling: Generates structures in common organic space groups.
• Force Field Assignment: Automated parameterization for organic molecules.
• Polymorph Screening: Identifies low-energy packing arrangements.
• Integration: Wraps PyXtal, RDKit, and optimization engines.

Inputs & Outputs

• Input formats: Molecular structure (SMILES, XYZ), search parameters.
• Output data types: Ranked crystal structures (CIF/POSCAR), energy tables.

Interfaces & Ecosystem

• PyXtal: Core engine for structure generation.
• RDKit: Molecule handling and conformation generation.
• LAMMPS/GULP: Classical optimization.
• VASP/QE: Ab initio optimization.

Verification & Sources

• Confidence: ✅ VERIFIED
• Primary Source: HTOCSP GitHub
• Reference: Paper associated with MaterSim group (Recent 2024/2025 work).