BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Exploring Chemical Space with Machine Learning
DTSTART:20231204T151500
DTEND:20231204T161500
DTSTAMP:20260501T085704Z
UID:348103da9dfc94aff0b0c2c7c84dd5824abb22e894c945032f5744bf
CATEGORIES:Conferences - Seminars
DESCRIPTION:Originally from Ukraine\, Ganna (Anya) Gryn’ova received her
  BS and MSc in chemistry summa cum laude from Oles Honchar Dnipro National
  University. In 2014 she received a PhD in computational chemistry from Au
 stralian National University. Her doctoral thesis gathered a number of awa
 rds\, including the IUPAC-Solvay International Award for Young Chemists fo
 r one of the five most outstanding PhD theses in the general area of the c
 hemical sciences worldwide. Dr. Gryn’ova continued her research career a
 t École Polytechnique Fédérale de Lausanne as a postdoctoral researc
 her working on in silico modeling of organic semiconductors. In 2016 she w
 on the Marie Skłodowska-Curie Actions individual fellowship and focussed 
 on the non-conventional architectures of single-molecule junctions. In 201
 9\, Dr. Gryn’ova started her independent scientific career leading the j
 unior research group “Computational Carbon Chemistry” (CCC) at the Hei
 delberg Institute for Theoretical Studies (HITS gGmbH) and Interdisciplina
 ry Center for Scientific Computing (IWR) at Heidelberg University\, German
 y. The CCC group uses state-of-the-art computational chemistry and data sc
 ience to explore and exploit diverse functional organic materials for appl
 ications in organocatalysis and environmental remediation. In 2021\, Anya 
 received the prestigious ERC Starting Grant for her project “PATTERNCHEM
 : Shape and Topology as Descriptors of Chemical and Physical Properties in
  Functional Organic Materials”\; she is also a principal investigator in
  the Collaborative Research Centre SFB1249 “N-Heteropolycycles as Functi
 onal Materials” and the SIMPLAIX strategic research initiative on bridgi
 ng scales from molecules to molecular materials by multiscale simulation a
 nd machine learning.\nChemical (molecular\, quantum) machine learning reli
 es on representing molecules in unique and informative ways. Here\, we int
 roduce two new representations – a quantum-inspired representation calle
 d matrix of orthogonalised atomic orbital coefficients (MAOC) [S. Llenga\,
  G. Gryn’ova\, J. Chem. Phys.\, 2023\, 158\, 214116]\, and a fragmentati
 on-based technique called matrix of fragment similarity representation (MF
 SR) [in prep.]. MAOC is uniquely suitable for representing monatomic\, mol
 ecular\, and periodic systems\, and can distinguish compounds with identic
 al compositions and geometries but distinct charges and spin multiplicitie
 s. MFSR is a dimensionality reduction and representation technique for map
 ping and exploring the chemical space based on specific building blocks. M
 ost industrially and biologically relevant macromolecules are formed as a 
 combination of finite building blocks (e.g.\, all proteins are a combinati
 on of just 20 aminoacids)\, and MFSR can predict their properties in less 
 than a fraction of a second and with the quantum-chemical accuracy. More b
 roadly\, MFSR can be applied to any chemical system using either atoms or 
 “ghost” centroids as representative fragments and allows even the most
  entangled deep learning models to be decodable in a chemically intuitive 
 form.
LOCATION:BCH 2218 https://plan.epfl.ch/?room==BCH%202218 https://epfl.zoom
 .us/j/68447908297?pwd=OU5JUGJUSUhZc0ZNYjQ2WENvYlNRdz09
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR