I thought of a #cheminformatics project idea to label the NMR chemical shift of atoms on a molecular structure, and it turns out that already exists: CASCADE, ChemicAl Shift CAlculation with DEep learning from Yanfei Guan and Robert Paton. Nice application of #ML #MachineLearning for #chemistry! Web interface at https://nova.chem.colostate.edu/cascade/predict/, code on GitHub https://github.com/patonlab/CASCADE. Example: 1H (proton) NMR shifts labeled on the repeat unit from my last blog post https://bertiewooster.github.io/2024/10/15/Color-from-Conjugation.html. #python #ML
=> More informations about this toot | More toots from jemonat@fosstodon.org
@jemonat just wondering, databases like NMRShiftDb have been doing that for a long time. is there something I am missing (besides things being easier to reuse nowadays)?
=> More informations about this toot | More toots from egonw@social.edu.nl
@egonw thanks for telling me about NMRShiftDb! CASCADE was the first tool I came upon. It does add machine learning so it may have some advantages.
=> More informations about this toot | More toots from jemonat@fosstodon.org
@jemonat NMRShiftDb does prediction too. if I remember correctly, the database was set up to collect enough training data? cc @steinbeck
=> More informations about this toot | More toots from egonw@social.edu.nl This content has been proxied by September (3851b).Proxy Information
text/gemini