Sharing data from NMR model prediction validation
Guillaume Godin
Giving computers a sense of smell | Principal Machine Learning Engineer @ Osmo | Data Science Director | Science Reviewer
Dear Community,
I am working on a NMR prediction 13C model build using NMRShiftDB2. The results look very good. Now I would like to get a more robust, independent and accurate dataset for a external validation aka (500-1000 molecules very well annotated and not in NMRShiftDB2). If you have such dataset, please contact me.
Best,
Guillaume
Director at ChemAlive SA
5 个月Cool. I agree with Zheng. Prediction with DFT. I could of course get involved to build that dataset. ;)
Perhaps train on DFT computed CS. Apply a scaling factor normalizing the CS from ab initio to existing dataset first, of course.
Founder, CEO & CSO @ AI | ffinity | NMR- & AI-driven Drug Design and Protein Structure Determination
5 个月Are small fragments suitable for your purpose?