BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20250712T230040EDT-1265l5WA5t@132.216.98.100
DTSTAMP:20250713T030040Z
DESCRIPTION:\n	Matrix completion in genetic methylation studies: LMCC\, a Li
 near Model of Coregionalization with informative Covariates\n\n	 \n\n\nAbst
 ract:\n\nDNA methylation is an important epigenetic mark that modulates ge
 ne expression through the inhibition of transcriptional proteins binding t
 o DNA. As in many other omics experiments\, missing values is an issue and
  appropriate imputation techniques are important to avoid an unnecessary s
 ample size reduction as well as to optimally leverage the information coll
 ected. We consider the case where a relatively small number of samples are
  processed via an expensive high-density Whole Genome Bisulfite Sequencing
  (WGBS) strategy and a larger number of samples are processed using more a
 ffordable low-density array-based technologies. In such cases\, one can im
 pute/complete the data matrix of the low coverage (array-based) methylatio
 n data using the high-density information provided by the WGBS samples. In
  this work\, we propose an efficient Linear Model of Coregionalization wit
 h informative Covariates (LMCC) to predict missing values based on observe
 d values and informative covariates. Our model assumes that at each genomi
 cs position\, the methylation vector of all samples is linked to the set o
 f fixed factors (covariates) and a set of latent factors. Furthermore\, we
  exploit the functional nature of the data and the spatial correlation acr
 oss positions/sites by assuming Gaussian processes on the fixed and latent
  coefficient vectors\, respectively. Our simulations show that the use of 
 covariates can significantly improve the accuracy of imputed values\, espe
 cially in cases where missing data contain some relevant information about
  the explanatory variable. We also show that the proposed model is efficie
 nt when the number of columns is much greater than the number of rows in t
 he data matrix-which is usually the case in methylation data analysis. Fin
 ally\, we apply and compare the proposed method with alternative approache
 s to complete a matrix of DNA methylation containing 15 rows (methylation 
 samples) and 1 million columns (sites). Joint work with Melina Ribaud and 
 Aurelie Labbe (HEC\, Montreal).\n\nSpeaker\n\nKarim Oualkacha is a profess
 or in the Department of Mathematics at Université du Québec à Montréal (UQ
 AM). He received BSc in Mathematics and MSc in Statistics and Operational 
 Research from Université Cadi Ayyad (Marrakech\, Morocco)\, and MSc and Ph
 D in Statistics from Université Laval (Quebec city). His research interest
 s focus on sparse multivariate statistical methods for high-dimensional da
 ta and modelling of multidimensional dependencies based on copulas\, with 
 applications in statistical genetics.\n\nhttps://mcgill.zoom.us/j/82678428
 848\n\nMeeting ID: 826 7842 8848\n\nPasscode: None\n\n \n
DTSTART:20240216T203000Z
DTEND:20240216T213000Z
LOCATION:Room 1104\, Burnside Hall\, CA\, QC\, Montreal\, H3A 0B9\, 805 rue
  Sherbrooke Ouest
SUMMARY:Karim Oualkacha (UQAM)
URL:/mathstat/channels/event/karim-oualkacha-uqam-3553
 50
END:VEVENT
END:VCALENDAR