| Latest release |   |
|---|---|
| Status |    |
| Community |  |
A playground for applying graph convolutional networks to molecules, with a focus on learning continuous "atom-type" embeddings and from these classical molecule force field parameters.
Note: This project is still in development and liable to substantial API and other changes.
This framework is mostly based upon the End-to-End Differentiable Molecular Mechanics Force Field Construction preprint by Wang, Fass and Chodera.
NAGL is bound by a Code of Conduct.
See our documentation for notes on installation, basic usage, theory, and examples!
The NAGL source code is hosted at https://github.com/openforcefield/openff-nagl and is available under the MIT license (see the file LICENSE). Some parts inherit from code distributed under other licenses, as detailed in LICENSE-3RD-PARTY).
NAGL inherits from Simon Boothroyd's NAGL library at https://github.com/SimonBoothroyd/nagl.