BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250822T115805Z
LOCATION:Campussaal - Plenary Room
DTSTART;TZID=Europe/Stockholm:20250616T102000
DTEND;TZID=Europe/Stockholm:20250616T105000
UID:submissions.pasc-conference.org_PASC25_sess149_pos149@linklings.com
SUMMARY:P07 - A Deep Dive into Deep Learning Frameworks for Protein Struct
 ure Prediction: Developing and Evaluating Classes of Biomolecular Complexe
 s
DESCRIPTION:Verónica G. Melesse Vergara, Érica Texeira Prates, Manesh Shah
 , and Dan Jacobson (Oak Ridge National Laboratory)\n\nAccurately predictin
 g the structure of a protein has been a long standing and extremely challe
 nging problem in biology. In recent years, the rapid evolution and adoptio
 n of artificial intelligence (AI) in scientific domains including biology 
 have made the prediction of protein structures leveraging deep learning (D
 L) frameworks with accuracy rivaling that of experimental crystal structur
 es possible. These advances are key to understanding protein function and 
 play a central role in accelerating the drug discovery process. \nThis wor
 k focuses on comparing the performance of state-of-the-art protein structu
 re prediction models across a predefined set of 7 challenging biomolecule 
 categories evaluated using AlphaFold2, AF2Complex, AlphaFold-multimer, Alp
 haFold3, Chai-1, and Boltz-1. The results compare the accuracy and perform
 ance of each method when applied to classes of challenging protein-protein
  complexes. The evaluation was conducted leveraging computational resource
 s at Oak Ridge National Laboratory (ORNL) including Frontier, ORNL’s exasc
 ale supercomputer. \nThe data set constructed, the resulting biomolecular 
 complex type classification, and  the comprehensive set of guidelines deri
 ved can aid in future experiment design. The results from this study can a
 lso provide a deeper understanding of the advantages and limitations of ea
 ch model when applied to specific classes of biomolecular complexes as opp
 osed to individual complexes.\n\nSession Chair: Chris Cantwell (Imperial C
 ollege London)\n\n
END:VEVENT
END:VCALENDAR