FlexCryst: A Force Field Based Suite for Computational Crystallography

A high-fidelity transformation hub for crystallographic and simulation data. It facilitates seamless format exchange between CIF, CSSR, MOL2, PDB, and DL_POLY (*.poly). This module acts as a critical scientific bridge between commercial suites like Materials Studio and molecular dynamics solvers.

When the Strict Filter is active, the engine applies AI-compliance "Cleanser" logic to verify lattice energy stability, stoichiometric consistency (Z'), and interatomic clash thresholds, ensuring only physically valid structures are used in downstream MD simulations or neural network training.

An advanced diagnostic visualization tool that provides a quantitative assessment of the crystal lattice. Beyond traditional rendering, it identifies all significant interatomic interactions—both stabilizing attractive contacts and destabilizing repulsions—based on statistical potentials. The module calculates precise distances and provides a breakdown of energy contributions, allowing for the immediate identification of structural anomalies in CIF, PDB, and RES files.

Utilizes a proprietary Similarity Index to enable the high-precision comparison of crystal structures. This algorithm goes beyond simple Cartesian RMSD by analyzing the coordination environment and symmetry-related motifs of the entire packing. It is an essential tool for clustering large structural ensembles, identifying polymorphs across different experimental sources, and quantifying the convergence of predicted models toward experimental targets.

The core analytical engine for lattice energy evaluation. It employs self-consistent empirical potentials derived through machine learning from the CSD. Score allows for the thermodynamic ranking of structural candidates and includes gradient-based geometry optimization. By relaxing atomic coordinates to their local minima within the empirical potential energy surface, it ensures that every structural model adheres to the geometric constraints observed in experimental crystallography.

Enables the ab initio prediction of crystal structures for rigid organic, organometallic, and inorganic compounds. Driven by original Data Mining Force Fields, the module derives potential parameters for all atom types directly from the CSD. It supports structure generation across the most common space groups and is capable of handling complex co-crystals (homo- and hetero-molecular structures) with up to 50 independent molecules in the asymmetric unit. By bridging molecular geometry and 3D packing, it provides a powerful platform for identifying potential polymorphs and screening for novel multi-component materials.

A specialized pre-processing module for X-ray powder diffraction (XRPD) data. It handles the critical task of transforming raw experimental scans into clean diffraction profiles. This includes automated background subtraction, line profile analysis, and peak fitting. By accounting for experimental broadening and noise, it prepares the diffraction data for accurate structural matching or full-pattern decomposition in the SDPD workflow.

Provides an automated solution for Structure Determination from Powder Diffraction (SDPD) data. By combining the FlexCryst similarity index with a global search algorithm, this module can identify the correct crystal packing even from un-indexed or low-resolution powder patterns. It simulates diffraction patterns for structural candidates and ranks them based on their fit to the experimental data, significantly accelerating the path from experimental scan to solved structure.

The ultimate tool for custom force field engineering. This module enables the refinement of empirical potential parameters for specific chemical domains. By mining targeted subsets of the CSD, researchers can optimize atom-type-specific potentials to handle unusual functional groups or novel molecular classes. This ensures that the Machine Learning potentials remain highly accurate even for chemistry that falls outside the standard general force field parameters.