BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:AI in chemistry and beyond: Machine learning for reactivity using 
 expert descriptors and mechanistic information
DTSTART:20230509T151500
DTEND:20230509T161500
DTSTAMP:20260528T020259Z
UID:47c49108a0242e3f797cb6593eb3c6a4430a3554c3baddfac7bc5a03
CATEGORIES:Conferences - Seminars
DESCRIPTION:Kjell Jorner is an Assistant Professor of Digital Chemistry at
  ETH Zurich since January 2023. His work focuses on accelerating chemical 
 discovery with digital tools\, with a special emphasis on reactivity and c
 atalysis. His group does interdisciplinary research\, drawing from the fie
 lds of computational chemistry\, cheminformatics and machine learning. Bef
 ore joining ETH Zurich\, he was a postdoctoral researcher with Alán Aspur
 u-Guzik (2021-2022) and at AstraZenecaUK (2018-2020). Kjell has a PhD from
  Uppsala University (2018) on computational physical organic chemistry for
  the photochemistry of aromatic compounds.\nDeep learning based on string 
 or graph representations of molecules has shown great progress in the last
  few years. Important applications include\, for example\, synthesis predi
 ction\, protein structure prediction and machine learning potentials. All 
 of these applications benefit from an abundance of well-curated datasets o
 n the order of at least hundreds of thousands of points. For many applicat
 ions in chemistry\, datasets are much smaller\, on the order of tens or hu
 ndreds of datapoints. Machine learning with classical methods has here bee
 n the gold standard\, based on expert-picked descriptors. These descriptor
 s are mostly problem-specific and often calculated with quantum-chemical s
 oftware. During the last few years\, we have developed the open-source Mor
 feus Python package for calculating descriptors\, mainly related to cataly
 sis and reactivity. Morfeus was for example used to calculate descriptors 
 for the Kraken database of phosphine ligand properties.While most reactivi
 ty models include only information on reactants and/or products\, increase
 d accuracy can be obtained by including information from high-energy inter
 mediates and transitions states along the reaction path. We will highlight
  our work on using mechanistic information to predict activation energies 
 and selectivities for the nucleophilic aromatic substitution reaction\, re
 presenting ~9% of all reactions carried out in the pharmaceutical industry
 .Although including mechanistic information can improve model performance\
 , obtaining the required high-energy structures is time-consuming and non-
 robust. We have therefore worked on robuster and faster methods based on f
 orce fields in the Polanyi package. Using Polanyi\, we recently created th
 e first reactivity benchmark task for generative models in the Tartarus su
 ite. Tartarus allows comparison on which are more chemically realistic and
  challenging than conventional tasks such as logP or QED optimization.
LOCATION:CH G1 495 https://plan.epfl.ch/?room==CH%20G1%20495
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR