BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Large-scale inference for detecting QTL hotspots in hierarchically
 -related sparse regression models
DTSTART:20211210T151500
DTEND:20211210T170000
DTSTAMP:20260408T071230Z
UID:d9279d9bb28717f1ddebe72367dea4602cf4c391af7f37aa2ccd85a7
CATEGORIES:Conferences - Seminars
DESCRIPTION:Hélène Ruffieux\, University of Cambridge\nThere is a broad 
 consensus in the biostatistical community about the merits of collectively
  accounting\nfor large numbers of genetic loci while leveraging informatio
 n across correlated outcomes to uncover shared\nregulation patterns. Howev
 er\, many also acknowledge the mathematical and computational diculties ha
 mpering\nthese practices in genetic association problems.\n\nHere we prese
 nt a scalable sparse regression framework for joint molecular quantitative
  trait locus (QTL)\nanalysis\, where associations between genetic variants
  and molecular gene products are sought. Speci cally\,\nwe devise two comp
 lementary approaches based on a series of parallel regressions combined in
  a hierarchical\nmanner to exibly accommodate high-dimensional responses (
 molecular outcomes) and predictors (genetic\nvariants)\, thereby allowing 
 information-sharing across outcomes and variants. We also directly model t
 he\npropensity of variants to be hotspots\, i.e.\, to remotely control the
  levels of many gene products\, via a dedicated\ntop-level representation.
  We implement variational inference procedures augmented with simulated an
 nealing\nschemes to enhance exploration of highly multimodal spaces and al
 low simultaneous analysis of thousands of\nsamples\, responses\, predictor
 s and predictor-level annotations. This uni ed learning boosts statistical
  power\nand helps shed light on the biological processes underlying geneti
 c regulation. We illustrate the advantages of\nour approaches in simulatio
 ns emulating real-data conditions and in a monocyte expression QTL study\,
  which\ncon rms known hotspots and reveals new ones\, as well as plausible
  mechanisms of action.\n 
LOCATION:MA B1 524 https://plan.epfl.ch/?room==MA%20B1%20524 https://epfl.
 zoom.us/j/65122471870
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR
