Institute of Crystallography - CNR

CrystalMELA

CrystalMELA (Crystallography MachinE LeArning) is a web-based platform that allows users to quickly and easily perfor automated crystallographic analysis from powder X-ray diffraction (PXRD) data using advanced machine learning (ML) methods.

The platform provides an intuitive interface that integrates multiple complementary ML models and supports the crystallographic task of:

  • Crystal system classification
  • Lattice parameter prediction (a, b, c, α, β, γ)

CrystalMELA processes PXRD profiles directly, without requiring peak detection or manual preprocessing. This enable a fast, flexible and user-friendly workflow that serves as an alternative to conventional crystallographic approaches. The platform is suitable for a wide range of polycrystalline materials spanning organic, inorganic, metal-organic and minerals. It can be used for both routine analyses and high-throughput studies, supporting the determination of the seven crystal systems and/or unit cell parameters.

RESEARCH AREA
KEYWORDS
References

CrystalMELA: a new crystallographic machine learning platform for crystal system determination
Nicola Corriero, Rosanna Rizzi, Gaetano Settembre, Nicoletta Del Buono and Domenico Diacono

DOI