BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Protein generation and fitness optimization - Jason Yim\, EECS\, M
 IT
DTSTART:20230620T160000Z
DTEND:20230620T170000Z
UID:TALK202555@talks.cam.ac.uk
CONTACT:Pietro Lio
DESCRIPTION:Deep learning is rapidly advancing the best computational tool
 s in computational biology. In this talk\, I will discuss recent advanceme
 nts in using generative models and sampling algorithms to generate protein
 s and optimize their fitness. First\, I will go over FrameDiff in which we
  lay the mathematical foundation of SE(3) diffusion and introduce a practi
 cal algorithm for training a frame-based generative model over protein bac
 kbones. SE(3) diffusion is then utilized in a state-of-the-art protein des
 ign method\, RFdiffusion\, that is pre-trained on protein structure predic
 tion. To conclude\, I will go over latest work in sequence-based protein f
 itness optimization using Gibbs with Gradients. We argue for the importanc
 e of model regularization for handling protein fitness sparsity and ensuri
 ng a smooth optimization landscape.\n\nhttps://cl-cam-ac-uk.zoom.us/j/9409
 4286505?pwd=T2V3MFVSQ1ZKTFFyYlFueHlzTE83Zz09\n\nBio: Jason Yim is a second
  year EECS PhD student at MIT advised by Tommi Jaakkola and Regina Barzila
 y. Previously\, he graduated from Johns Hopkins University with a bachelor
 s in computer science then worked at DeepMind at a research engineer on me
 dical imaging and AlphaFold-multimer. His research interest is to develop 
 machine learning methods with applications to scientific problems. His lat
 est works focus on inverse problems in protein design.
LOCATION:Lecture Theatre 2\, Computer Laboratory\, William Gates Building 
 and Zoom
END:VEVENT
END:VCALENDAR