NMR Parameter Prediction with Machine Learning

  • William Gerrard

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

The prediction of NMR parameters through machine learning was investigated and several highly accurate prediction algorithms developed. Prediction models sensitive to 3-Dimensional structure in small molecules are presented for chemical shifts and scalar coupling constants, several of which outperform current state-of-the-art algorithms. Several large, high quality DFT datasets were also produced, their construction and composition are detailed in this work. Finally the application of the newly developed prediction algorithms to a realistic diastereomer discrimination task is explored, along with the adaptation of one of the machine learning frameworks to the prediction of binding affinities.
Date of Award2 Dec 2021
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorAdrian J Mulholland (Supervisor) & Craig P Butts (Supervisor)

Cite this

'