BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Using artificial intelligence to push the limits of one-dimensiona
 l NMR for structure elucidation  - Professor Thomas E. Markland\, Stanford
  University
DTSTART:20260121T143000Z
DTEND:20260121T153000Z
UID:TALK231946@talks.cam.ac.uk
CONTACT:Lisa Masters
DESCRIPTION:Rapid determination of molecular structures can greatly accele
 rate workflows across many chemical disciplines. However\, elucidating str
 ucture using only one-dimensional (1D) NMR spectra\, one of the most widel
 y used techniques for the characterization of organic compounds and natura
 l products\, remains an extremely challenging problem because of the combi
 natorial explosion of the number of possible molecules as the number of co
 nstituent atoms is increased – for example for molecules with up to 36 n
 on-hydrogen atoms\, the number of possible structures consistent with the 
 typical bonding rules of chemistry has been estimated to range from 1020-1
 060. The task of determining the molecular structure (formula and connecti
 vity) of a molecule of this size using only its one-dimensional 1H and/or 
 13C NMR spectrum\, i.e.\, de novo structure generation\, thus appears comp
 letely intractable. I will show how it is possible to achieve this task fo
 r systems with up to 40 non-hydrogen atoms across the full elemental cover
 age typically encountered in organic chemistry (C\, N\, O\, H\, P\, S\, Si
 \, B\, and the halogens) using a deep learning framework\, thus covering a
  vast portion of the drug-like chemical space. Leveraging insights from na
 tural language processing\, I will show how our transformer-based architec
 ture can overcome the combinatorial growth of the chemical space while als
 o being extensible to experimental data via fine-tuning.
LOCATION:Unilever Lecture Theatre\, Yusuf Hamied Department of Chemistry
END:VEVENT
END:VCALENDAR