BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20250330T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251026T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260425T051708Z
UID:1745832600@ist.ac.at
DTSTART:20250428T113000
DTEND:20250428T123000
DESCRIPTION:Speaker: Renana Gershoni-Poranne\nhosted by Alexander Bronstein
  \nAbstract: Polycyclic aromatic systems (PASs) present a seemingly insurm
 ountable challenge: vast chemical spaces\, complex electronic structures\,
  and elusive aromatic properties. Our mission\, should we choose to accept
  it\, is to harness the power of deep learning to decode these molecular m
 ysteries. In this talk\, we embark on a journey through this complex chemi
 cal space\, combining traditional computational methods with cutting-edge 
 artificial intelligence tools. We demonstrate how neural networks can be t
 rained to predict electronic properties with unprecedented speed and accur
 acy. More importantly\, we show how they can be used interpretably to extr
 act chemical insight.To enable the application of such techniques to the d
 esign of novel functional PASs\, we established the COMPAS Project -- a CO
 Mputational database of Polycyclic Aromatic Systems – which already cont
 ains ~1 million molecules in three datasets.1–4 We also developed two ne
 w types of molecular representation to enable efficient and effective mach
 ine- and deep-learning models to train on the new data: a) a text-based re
 presentation5 and b) a graph-based representation.6 By analyzing thousands
  of PAS structures with our dedicated representations\, our AI agents not 
 only achieve higher predictive ability with fewer data but have also uncov
 ered hidden patterns and structure-property relationships that traditional
  methods might have missed.Finally\, we implemented the first guided diffu
 sed-based model for inverse design of PASs: GaUDI.7 Our model generates ne
 w PASs with defined target properties. In addition to its flexible target 
 function and high validity scores\, GaUDI also accomplishes design of mole
 cules with properties beyond the distribution of the training data.From sm
 all benzenoid systems to extended graphene-like structures\, we show how d
 eep learning can navigate this complex chemical landscape\, offering new i
 nsights into molecular design and property prediction. This talk will not 
 self-destruct in five seconds\, but it will revolutionize how we think abo
 ut combining artificial intelligence with molecular science.    (1) 
           Wahab\, A.\; Pfuderer\, L.\; Paenurk\, E.\; Gershoni-Po
 ranne\, R. The COMPAS Project: A Computational Database of Polycyclic Arom
 atic Systems. Phase 1: Cata-Condensed Polybenzenoid Hydrocarbons. J. Chem.
  Inf. Model. 2022\, 62 (16)\, 3704. https://doi.org/10.1021/acs.jcim.2c005
 03.(2)           Mayo Yanes\, E.\; Chakraborty\, S.\; Gershoni-P
 oranne\, R. COMPAS-2: A Dataset of Cata-Condensed Hetero-Polycyclic Aromat
 ic Systems. Sci. Data 2024\, 11 (1)\, 97. https://doi.org/10.1038/s41597-0
 24-02927-8.(3)           Chakraborty\, S.\; Yanes\, E. M.\; Gers
 honi-Poranne\, R. Hetero-Polycyclic Aromatic Systems: A Data-Driven Invest
 igation of Structure–Property Relationships. Beilstein J. Org. Chem. 202
 4\, 20 (1)\, 1817–1830. https://doi.org/10.3762/bjoc.20.160.(4)    
        Wahab\, A.\; Gershoni-Poranne\, R. COMPAS-3: A Dataset of Per
 i-Condensed Polybenzenoid Hydrocarbons. Phys. Chem. Chem. Phys. 2024\, 26 
 (21)\, 15344–15357. https://doi.org/10.1039/D4CP01027B.(5)       
     Fite\, S.\; Wahab\, A.\; Paenurk\, E.\; Gross\, Z.\; Gershoni-Poran
 ne\, R. Text-Based Representations with Interpretable Machine Learning Rev
 eal Structure-Property Relationships of Polybenzenoid Hydrocarbons. J. Phy
 s. Org. Chem. 2022\, e4458. https://doi.org/10.1002/poc.4458.(6)     
       Weiss\, T.\; Wahab\, A.\; Bronstein\, A. M.\; Gershoni-Poranne\
 , R. Interpretable Deep-Learning Unveils Structure–Property Relationship
 s in Polybenzenoid Hydrocarbons. J. Org. Chem. 2023\, 88 (14)\, 9645–965
 6. https://doi.org/10.1021/acs.joc.2c02381.(7)           Weiss\,
  T.\; Mayo Yanes\, E.\; Chakraborty\, S.\; Cosmo\, L.\; Bronstein\, A. M.\
 ; Gershoni-Poranne\, R. Guided Diffusion for Inverse Molecular Design. Nat
 . Comput. Sci. 2023\, 3 (10)\, 873–882. https://doi.org/10.1038/s43588-0
 23-00532-0.
LOCATION:Raiffeisen Lecture Hall\, ISTA
ORGANIZER:Diana.Gruber@ista.ac.at
SUMMARY:Renana Gershoni-Poranne: Mission ImPASsible: Decoding Polycyclic Ar
 omatic Systems with Deep Learning
URL:https://talks-calendar.ista.ac.at/events/5614
END:VEVENT
END:VCALENDAR