BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
 M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20240418T112906Z
DTSTART:20240422T130000Z
DTEND:20240422T140000Z
SUMMARY:Gabriele Schweikert -- Computational Strategies to Analyse\, Impu
 te and Interpret Epigenetic Control Mechanisms of Gene Regulation [IN PE
 RSON]
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}w2q3-ls05hr
 ld-xj60b5
DESCRIPTION:Join us for this seminar by Gabriele Schweikert (Dundee) as p
 art of the North West Seminar Series in Mathematical Biology and Data Sc
 iences.\n\nThe talk will be hosted in person in room 4.63 of the Simon B
 uilding. For those who cannot attend in person the talk will also be str
 eamed via zoom\, please contact carl.whitfield@manchester.ac.uk or igor.
 chernyavsky@manchester.ac.uk for the zoom link\, or sign up to the maili
 ng list.\n\nTitle: Computational Strategies to Analyse\, Impute and Inte
 rpret Epigenetic Control Mechanisms of Gene Regulation\n\nAbstract: Epig
 enomic modifications are reversible chemical marks on top of the DNA\, t
 hat do not change the underlying sequence itself. Personal\, cell-type-s
 pecific epigenomes result from a combination of genetic variants and a c
 ellular memory of past cellular events. Epigenetic mechanisms are theref
 ore essential mediators of gene–environment interactions. Functionally\,
  they contribute to the control of current and future transcription and 
 thus play important roles during development\, disease progression and a
 geing. Recently\, efforts to record personal epigenomes across tissues h
 ave become feasible. However\, the large number of assays required for a
  complete epigenomic map continues to be a limiting factor for personali
 sed epigenomics.\n\nMachine learning approaches are poised to fill this 
 gap. In this talk I will present eDICE\, which is based on the transform
 er architecture and is capable of predicting individual-specific epigeno
 mic landscapes. We achieve high prediction accuracy by learning factoris
 ed representations. At the same time\, eDICE has unprecedented generalis
 ation capabilities. The complete model fits into GPU memory and does not
  require complicated training schemes as the number of parameters is sev
 eral orders of magnitude smaller than in previous models. These are esse
 ntial preconditions to apply computational imputation for personalised e
 pigenomics and to use these methods at scale.\n\nTo subscribe to the mai
 ling list for this event series\, please send an e-mail with the phrase 
 “subscribe math-lifesci-seminar” in the message body to listserv@listser
 v.manchester.ac.uk
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
LOCATION:4.63\, Simon Building\, Manchester
END:VEVENT
END:VCALENDAR