BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:MOFA: a principled framework for the unsupervised integration of m
 ulti-omics data - Ricard Argelaguet
DTSTART:20200120T163000Z
DTEND:20200120T170000Z
UID:TALK137476@talks.cam.ac.uk
CONTACT:Bingqing Cheng
DESCRIPTION:The emergence of high-throughput technologies and the increasi
 ng availability of clinical data are radically changing the study of biolo
 gy and its medical applications. In particular\, the profiling of multiple
  molecular layers (omics) from the same patient\, provides a unique opport
 unity to build statistical models to understand the molecular sources of p
 atient heterogeneity. \nI will present MOFA\, a matrix factorisation frame
 work for the comprehensive integration of multi-omics data. MOFA builds up
 on a Group Factor Analysis framework combined with fast variational Bayes 
 inference. The model pools information across all -omics to reconstruct a 
 low-dimensional representation of the data\, thereby enhancing data interp
 retation and facilitating the definition of predictive models for clinical
  outcomes.\nTo demonstrate its practical utility\, I will present an appli
 cation of MOFA on a cohort of 200 patient samples of chronic lymphocytic l
 eukaemia that were profiled using multiple molecular assays\, including so
 matic mutations\, RNA expression\, DNA methylation and ex vivo drug respon
 ses. MOFA identified major dimensions of disease heterogeneity\, including
  mutations on the immunoglobulin heavy-chain variable region and trisomy o
 f chromosome 12.
LOCATION:Mott Seminar (531) room\, top floor of the Mott Building\, in the
  Cavendish Laboratory\, West Cambridge.
END:VEVENT
END:VCALENDAR