Software

1. PolymerGNN

PolymerGNN is a deep learning method to predict polymer properties. This model utilizes state-of-the-art graph neural networks (GNNs), which encode molecular graphs into fixed-length vectors.

• Code
• Publication: O. Queen, G. A. McCarver, S. Thatigotla, B. P. Abolins, C. L. Brown, V. Maroulas, K. D. Vogiatzis Polymer Graph Neural Networks for Multitask Property Learning, npj Comput. Mater., 2023, 9, 90.

2. Molecular Representations

Hands-on examples that accompany our eBook on molecular representations for machine learning applications.

• Code
• Publication: G. M. Jones, B. Story, V. Maroulas, K. D. Vogiatzis Molecular Representations for Machine Learning, ACS In Focus Series, 2023.

3. Persistent Images

Persistent images (PIs) are a molecular representations for chemical machine learning applications based on persistent homology, an applied branch of topology. 

• Code
• Publication: J. Townsend, C. P. Micucci, J. H. Hymel, V. Maroulas, K. D. Vogiatzis, Representation of molecular structures with persistent homology for machine learning applications in chemistry, Nat. Commun., 2020, 11, 3230.

Application of PIs on a database with ~181k Fe(IV)-oxo complexes + Similarity comparison based on PIs

• Code
• Publication: G. M. Jones, B. A. Smith, J. K. Kirkland, K. D. Vogiatzis Data-Driven Ligand Field Exploration of Fe(IV)-oxo Sites for C-H Activation, Inorg. Chem. Front., 2023, 10, 1062.

4. Data-driven Quantum Chemistry

Hands-on examples on data-driven coupled-cluster (DDCC) and data-driven complete active space second-order perturbation theory (DDCASPT2).

• Code
• Publication: G. M. Jones, P. D. V. S. Pathirage, K. D. Vogiatzis, Data-driven Acceleration of Coupled-Cluster Theory and Perturbation Theory Methods, in: “Quantum Chemistry in the Age of Machine Learning”, 2022, Editor: Pavlo Dral, Elsevier, pp. 509-529.

5. Data-driven v2RDM

Repository of the data-driven variational 2RDM (v2RDM) project with an example code.

• Code
• Publication: G. M. Jones, R. R. Li, A. E. DePrince III, K. D. Vogiatzis Data-driven Refinement of Electronic Energies from Two-electron Reduced-density-matrix Theory, J. Phys. Chem. Lett., 2023, 14, 6377.