Open ml-evs opened 12 months ago
This is now available (temporarily) at https://optimade-gnome.odbx.science/v1/structures (code at https://github.com/ml-evs/optimade-gnome), if anyone is interested. Won't register this API with the OPTIMADE federation (yet) to avoid any future clashes pending the response to this issue. This means you can shoot any OPTIMADE query at it and filter by composition, formula etc.
Hi! Thanks for reaching out
There is on-going work to add this data to the Materials Project. My understanding is their integration will directly lead to the materials being available in OPTIMADE? If so, should we hold off so there are not duplicate entries?
Let me know what you think, happy to chat more :)
Duplicate entries are no problem (there is already huge duplication where e.g., NOMAD has ingested all other databases, and there's obviously overlap anyway -- the OPTIMADE model is of a decentralized federation of providers not one gigantic consistent database). It's up to the MP then really, and how up-to-date their own OPTIMADE API is. I guess if they are recomputing everything with their own workflows this may take a while? I'll just leave my server running in the meantime as I'm finding it useful to explore as-is!
Happy to discuss other options further in a call -- feel free to close this issue whenever you see fit.
Sounds great! OPTIMADE looks super cool and useful for materials community.
Let's chat more :) Would also love to get your initial impressions on the dataset. Feel free to email me at amilmerchant@google.com and we can set up some time to discuss.
I'll leave this issue open. If we get a lot of +1s or interest, we can make this a priority. 💯
Thanks for releasing this great dataset. Is there any interest in making it available with an OPTIMADE API (https://optimade.org) and joining our existing federation of 20 or so materials data providers? This will allow for unified queries for stable materials across the unified hull from e.g. MP, OQMD, Alexandria, MCloud, matterverse and others, which we believe will also enable other labs to more easily screen your proposals.
We have tools that make this process relatively straightforward at https://github.com/Materials-Consortia/optimade-python-tools
While it would be nice to have it officially hosted (and dynamic as you discover more!) if you are interested, I can also do this myself in the meantime (either on our servers or on the Materials Cloud).