English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Despite its fundamental importance and widespread use for assessing reactionsuccess in organic chemistry, deducing chemical structures from nuclearmagnetic resonance (NMR) measurements has remained largely manual and timeconsuming. To keep up with the accelerated pace of automated synthesis in selfdriving laboratory settings, robust computational algorithms are needed torapidly perform structure elucidations. We analyse the effectiveness of solvingthe NMR spectra matching task encountered in this inverse structure elucidationproblem by systematically constraining the chemical search space, andcorrespondingly reducing the ambiguity of the matching task. Numerical evidencecollected for the twenty most common stoichiometries in the QM9-NMR data baseindicate systematic trends of more permissible machine learning predictionerrors in constrained search spaces. Results suggest that compounds withmultiple heteroatoms are harder to characterize than others. Extending QM9 by$\sim$10 times more constitutional isomers with 3D structures generated bySurge, ETKDG and CREST, we used ML models of chemical shifts trained on theQM9-NMR data to test the spectra matching algorithms. Combining both$^{13}\mathrm{C}$ and $^{1}\mathrm{H}$ shifts in the matching process suggeststwice as permissible machine learning prediction errors than for matching basedon $^{13}\mathrm{C}$ shifts alone. Performance curves demonstrate that reducingambiguity and search space can decrease machine learning training data needs byorders of magnitude.

Impact of noise on inverse design: The case of NMR spectra matching