DScribe: Library of descriptors for machine learning in materials science

2019/09/26 by Lauri Himanen, Marc O. J. Jäger, Marc O.J. Jäger +7 · 822 citations
Chemistry · Computer Science · Materials Science · Mathematics · #Algorithm #Artificial intelligence #Chemistry #Computational Drug Discovery Methods #Computational science #Computer science #Feature (linguistics) #Geometry #Implementation #Machine Learning in Materials Science #Mathematics #Matrix (chemical analysis) #Matrix multiplication #Physics #Programming language #Quantum mechanics #Software #Tensor (intrinsic definition) #X-ray Diffraction in Crystallography

paper · pdf · doi:10.1016/j.cpc.2019.106949

published in Computer Physics Communications 247, 106949 (Elsevier BV)

openalex publication_date 2019/09/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

DScribe is a software package for machine learning that provides popular feature transformations (“descriptors”) for atomistic materials simulations. DScribe accelerates the application of machine learning for atomistic property prediction by providing user-friendly, off-the-shelf descriptor implementations. The package currently contains implementations for Coulomb matrix, Ewald sum matrix, sine matrix, Many-body Tensor Representation (MBTR), Atom-centered Symmetry Function (ACSF) and Smooth Overlap of Atomic Positions (SOAP). Usage of the package is illustrated for two different applications: formation energy prediction for solids and ionic charge prediction for atoms in organic molecules. The package is freely available under the open-source Apache License 2.0. Program Title: DScribe Program Files doi: 10.17632/vzrs8n8pk6.1 Licensing provisions: Apache-2.0 Programming language: Python/C/C++ Supplementary material: Supplementary Information as PDF Nature of problem: The application of machine learning for materials science is hindered by the lack of consistent software implementations for feature transformations. These feature transformations, also called descriptors, are a key step in building machine learning models for property prediction in materials science. Solution method: We have developed a library for creating common descriptors used in machine learning applied to materials science. We provide an implementation the following descriptors: Coulomb matrix, Ewald sum matrix, sine matrix, Many-body Tensor Representation (MBTR), Atom-centered Symmetry Functions (ACSF) and Smooth Overlap of Atomic Positions (SOAP). The library has a python interface with computationally intensive routines written in C or C++. The source code, tutorials and documentation are provided online. A continuous integration mechanism is set up to automatically run a series of regression tests and check code coverage when the codebase is updated.

Citations

• Die Berechnung optischer und elektrostatischer Gitterpotentiale
• Quantum-chemical insights from deep tensor neural networks

Cited by

• NepTrain and NepTrainKit: Automated active learning and visualization toolkit for neuroevolution potentials
• q-pac : A Python package for machine learned charge equilibration models
• Conformational dynamics of the pyrene excimer
• ULaMDyn: enhancing excited-state dynamics analysis through streamlined unsupervised learning
• Neural Network Potentials: A Concise Overview of Methods
• Identifying phase transitions in zeolitic imidazolate frameworks: microscopic insight from molecular simulations
• Alchemical harmonic approximation based potential for iso-electronic diatomics: Foundational baseline for Δ-machine learning
• Investigating 3D Atomic Environments for Enhanced QSAR
• Reproducible container solutions for codes and workflows in materials science
• Generalization of Long-Range Machine Learning Potentials in Complex Chemical Spaces
• Characterizing Defect Dynamics in Silicon Carbide Using Symmetry-Adapted Collective Variables and Machine Learning Interatomic Potentials
• Machine-Learned Interatomic Potentials for Structural and Defect Properties of YBa₂Cu₃O_7-δ
• Benchmarking GNNs for OOD Materials Property Prediction with Uncertainty Quantification
• Global geometry of chemical graph neural network representations in terms of chemical moieties
• Four Generations of High-Dimensional Neural Network Potentials
• Unsupervised Machine-Learning Pipeline for Data-Driven Defect Detection and Characterisation: Application to Displacement Cascades
• Gaussian Process Regression for Materials and Molecules
• DeFecT-FF: Accelerated Modeling of Defects in Cd-Zn--Te-Se-S Compounds Combining High-Throughput DFT and Machine Learning Force Fields
• dynsight: an Open Python Platform for Simulation and Experimental Trajectory Data Analysis
• Kinetics of Peierls dimerization transition: Machine learning force-field approach
• Spectral Analysis of Molecular Features: When Richer Features Do Not Guarantee Better Generalization
• Utilizing SciPy and other open source packages to provide a powerful API for materials manipulation in the Schrödinger Materials Suite
• Improved capabilities of the TurboGAP code for radiation induced cascade simulations: an illustration with silicon
• RAFFLE: active learning accelerated interface structure prediction
• Scalable Machine Learning Model for Energy Decomposition Analysis in Aqueous Systems
• PASS: Perturbation augmented space group structure sampling for transferable Fe-O machine learning interatomic potential
• Artificial intelligence for energy materials research: From classical machine learning to large models
• El Agente Sólido: A New Age(nt) for Solid State Simulations
• Ligand design for 227 Ac extraction by active learning and molecular topology
• Substituting density functional theory in reaction barrier calculations for hydrogen atom transfer in proteins
• A Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Medium-range structural order in amorphous arsenic
• LeMat-Traj: A Scalable and Unified Dataset of Materials Trajectories for Atomistic Modeling
• Metatensor and metatomic: foundational libraries for interoperable atomistic machine learning
• Recent Advances and Applications of Deep Learning Methods in Materials Science
• Chemical Space of Molecular Nanomotors: Optimizing Photochemical Properties for One- and Two-photon Applications
• Data-efficient machine-learning of complex Fe-Mo intermetallics using domain knowledge of chemistry and crystallography
• Atomistic Generative Diffusion for Materials Modeling
• Automatic generation of input files with optimised k -point meshes for Quantum ESPRESSO self-consistent field single-point total energy calculations
• Leveraging active learning-enhanced machine-learned interatomic potential for efficient infrared spectra prediction
• Efficient dataset generation for machine learning perovskite alloys
• Particle Swarm Based Hyper-Parameter Optimization for Machine Learned\n Interatomic Potentials
• NepTrain and NepTrainKit: Automated Active Learning and Visualization Toolkit for Neuroevolution Potentials
• Machine-Learned Potentials for Solvation Modeling
• Interpretable machine learned predictions of adsorption energies at the metal--oxide interface
• Kernel Ridge Regression for conformer ensembles made easy with Structured Orthogonal Random Features
• Predicting electronic screening for fast Koopmans spectral functional calculations
• From oligomers to entangled polymers: How to train a transferable machine learning interatomic potential
• Quotient Complex Transformer (QCformer) for Perovskite Data Analysis
• Graph Neural Network Predictions of Carbon 1s Binding Energies with Near-Experimental Accuracy
• Transition States Energies from Machine Learning: An Application to Reverse Water-Gas Shift on Single-Atom Alloys
• On phase separation and crystallization of Ge-rich GeSbTe alloys from atomistic simulations with a machine learning interatomic potential
• Hybrid localized graph kernel for machine learning energy-related\n properties of molecules and solids
• Global optimization of atomic structures with gradient-enhanced Gaussian process regression
• A Physically-informed Graph-based Order Parameter for the Universal Characterization of Atomic Structures
• Deep learning model for finding new superconductors
• Machine-Learning-Assisted De Novo Design of Organic Molecules and Polymers: Opportunities and Challenges. [europepmc]
• Digital navigation of energy-structure-function maps for hydrogen-bonded porous molecular crystals. [europepmc]
• Machine Learning Force Fields. [europepmc]
• How Machine Learning Will Revolutionize Electrochemical Sciences. [europepmc]
• Active discovery of organic semiconductors. [europepmc]
• Machine Learning Meets with Metal Organic Frameworks for Gas Storage and Separation. [europepmc]
• Unsupervised Learning Methods for Molecular Simulation Data. [europepmc]
• Geometric landscapes for material discovery within energy-structure-function maps. [europepmc]
• MLatom 2: An Integrative Platform for Atomistic Machine Learning. [europepmc]
• A Data-Driven Dimensionality Reduction Approach to Compare and Classify Lipid Force Fields. [europepmc]
• Gaussian Process Regression for Materials and Molecules. [europepmc]
• Differentiable sampling of molecular geometries with uncertainty-based adversarial attacks. [europepmc]
• Controlling Exchange Pathways in Dynamic Supramolecular Polymers by Controlling Defects. [europepmc]
• A transferable active-learning strategy for reactive molecular force fields. [europepmc]
• Choosing the right molecular machine learning potential. [europepmc]
• Learning in continuous action space for developing high dimensional potential energy models. [europepmc]
• Predicting the emission wavelength of organic molecules using a combinatorial QSAR and machine learning approach. [europepmc]
• Analogy Powered by Prediction and Structural Invariants: Computationally Led Discovery of a Mesoporous Hydrogen-Bonded Organic Cage Crystal. [europepmc]
• Classifying soft self-assembled materials via unsupervised machine learning of defects. [europepmc]
• MOFormer: Self-Supervised Transformer Model for Metal-Organic Framework Property Prediction. [europepmc]
• Computational Tools for Handling Molecular Clusters: Configurational Sampling, Storage, Analysis, and Machine Learning. [europepmc]
• MLatom 3: A Platform for Machine Learning-Enhanced Computational Chemistry Simulations and Workflows. [europepmc]
• Machine learning interatomic potential: Bridge the gap between small-scale models and realistic device-scale simulations. [europepmc]
• Self-Driving Laboratories for Chemistry and Materials Science. [europepmc]
• The design space of E(3)-equivariant atom-centred interatomic potentials. [europepmc]

Related

• Descriptors for Machine Learning of Materials Data
• Obtaining transferable chemical insight from solving machine-learning classification problems: Thermodynamical properties prediction, atomic composition as good as Coulomb matrix
• Beyond potentials: integrated machine-learning models for materials
• Atoms as Words: A Novel Approach to Deciphering Material Properties using NLP-inspired Machine Learning on Crystallographic Information Files (CIFs)
• Atom-Density Representations for Machine Learning