BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:NMR Prediction Uncertainty Enables DFT-Free Structural Confirmatio
 n - Ruslan Kotlyarov\, University of Cambridge
DTSTART:20250129T143000Z
DTEND:20250129T150000Z
UID:TALK226738@talks.cam.ac.uk
CONTACT:Lisa Masters
DESCRIPTION:While density functional theory (DFT) remains the standard for
  accurate simulation of nuclear magnetic resonance (NMR) spectra\, its com
 putational cost remains prohibitive. Use of DFT for structural confirmatio
 n is only justified where it offers substantial time savings over the expe
 riment\, such as total synthesis of natural products. Neural networks are 
 a promising solution for simpler molecules\, but published examples cannot
  estimate the prediction uncertainty. \n \nBy incorporating uncertainty es
 timation into an existing neural network\, we can confirm the structure fr
 om its NMR spectrum 100\,000 times faster than using DFT\, with calculatio
 ns completed in milliseconds rather than hours. Large-scale combinatorial 
 studies show that our approach matches accuracy of DFT-based DP5 analysis 
 and exceeds the sensitivity of simple error analysis. Analysis of 24 misas
 signed natural product structures demonstrates the generalisability of the
  method and equal performance to that of DFT.\n \nWe are now exploring the
  potential of the new method for automated structure revision and interpre
 tation of 1H NMR spectra.
LOCATION:Unilever Lecture Theatre\, Yusuf Hamied Department of Chemistry
END:VEVENT
END:VCALENDAR