BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:CIS - "Get to know your neighbors" Seminar series: Dr. Wouter Sael
 ens - Deplancke Lab
DTSTART:20230306T151500
DTEND:20230306T161500
DTSTAMP:20260427T224558Z
UID:1d80659b479f3dc932e34ffe7f96463abd0159b7d51a28be2ce5ac4e
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr. Wouter Saelens\nTitle: “Mountains in our DNA: how combi
 ning probabilistic modeling with neural networks deepens our understanding
  of omics data and genetic variation“\n\nAbstract: The move towards pre
 cision medicine will require a detailed and multiscale understanding of th
 e various molecular layers of the human cell. While progress in miniaturiz
 ation and sequencing has now enabled increasingly sensitive and high-resol
 ution profiling of the cell\, these have presented some unique machine-lea
 rning and statistical challenges that are far from resolved. It is clear t
 hat any molecular or cellular process is incredibly complex\, with numerou
 s non-linearities and interactions\, which would therefore require (deep) 
 neural networks to properly fit and predict the underlying biology. Howeve
 r\, limited sample sizes and lack of interpretability often mean we have t
 o rely on simplified\, but statistically convenient\, models that tend to 
 trade accuracy with generalizability.\n\nHere\, I will show in two example
 s how these statistically-convenient models can easily hide the true biolo
 gy. I first highlight how DNA accessibility data contains much more inform
 ation than previously thought\, and how this information can be extracted 
 by combining the expressivity of neural networks and normalizing flows\, w
 ith the statistical power of Bayesian inference. This approach not only de
 epens our understanding of cellular behavior but most importantly improves
  our predictive power for genetic variations that underlie diseases. Next\
 , I show how a similar approach can be applied on the cellular level to st
 udy the multifactorial effects of genetic perturbations on the cellular ph
 enotype. We propose that a modular\, interpretable but expressive\, framew
 ork is needed that can convert the various molecular and cellular features
  of a cell with a flexible probabilistic model.\n\nBio: Wouter Saelens is
  a postdoc in the lab of prof. Bart Deplancke\, where he studies the inter
 action within and between cellular processes (differentiation\, cell cycle
 \, ) and the several molecular layers (transcriptome\, proteome and chroma
 tin accessibility). Molecular biologist by training\, Wouter started combi
 ning his love for biology with machine learning during his PhD at VIB-Ghen
 t University in Ghent\, Belgium.\n\n\nThe CIS seminar will take place In h
 ybrid mode: Room INF 328 and by Zoom (Link TBD)\n\nPlease connect to your 
 zoom account using your "@epfl.ch" address\, as this live event is only op
 en to the EPFL community\nMonday\, March 6\, 2023 from 3:15 to 4:15 pm\nNB
 : Video recordings of the seminars will be made available on our website
  and published on our social media pages
LOCATION:INF 328 https://plan.epfl.ch/?room==INF%20328 https://epfl.zoom.u
 s/j/64775322490
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR