⚠️ Completely unofficial and separate from the deepmind work, use at your own risk. You should consult the data license on the deepmind repo for re-use limitations (CC-BY NC).⚠️
Taking some data from google-deepmind/materials_discovery and making it available as an OPTIMADE API, at least temporarily, at https://optimade-gnome.odbx.science. This only contains the PBE dataset for now.
The motivation is that this dataset can then be easily filtered by property and composition, returning structures that can be used by other labs (autonomous or otherwise), and also serendipitously, without needing to know which database to search in ahead of time.
Uses optimade-python-tools and optimade-maker.
- Clone with submodule.
- Download data using provided scripts in submodule, after
pip install -r requirements.txtfrom this repo. python materials_discovery/scripts/download_data_wget.py - Install the currently named archive-optimade-integration for easy ingestion
directly from static files, using the
optimade.yamlfile. - Use the
optimakeCLI to create an OPTIMADE-compliant JSONL file, then insert into your db of choice and host it wherever you like...
- Filter out any phases that include lathanides or actinides with
?filter=NOT elements HAS ANY "Ac","Th","Pa","U","Np","Pu","Am","Cm","Bk","Cf","Es","Fm","Md","No","La","Ce","Pr","Nd","Pm","Sm","Eu","Gd","Tb","Dy","Ho","Er","Tm","Yb"(leaving around 120k phases) - Filter only for binary and ternary phases with
?filter=nelements > 1 AND nelements < 4 - Binary phases excluding the lanthanides/actinides:
?filter=nelements=2 AND NOT elements HAS ANY "Ac", "Th", "Pa", "U", "Np", "Pu", "Am", "Cm", "Bk", "Cf", "Es", "Fm", "Md", "No", "La", "Ce", "Pr", "Nd", "Pm", "Sm", "Eu", "Gd", "Tb", "Dy", "Ho", "Er", "Tm", "Yb"
Paper with the source data:
Merchant, A., Batzner, S., Schoenholz, S.S. et al. Scaling deep learning for materials discovery. Nature (2023). https://doi.org/10.1038/s41586-023-06735-9
OPTIMADE format/federation description:
Andersen, C.W., Armiento, R., Blokhin, E. et al. OPTIMADE, an API for exchanging materials data. Sci Data 8, 217 (2021). https://doi.org/10.1038/s41597-021-00974-z
Tools used:
Evans et al., (2021). optimade-python-tools: a Python library for serving and consuming materials data via OPTIMADE APIs. Journal of Open Source Software, 6(65), 3458, https://doi.org/10.21105/joss.03458