BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Taking QTL mapping in multiparental populations to the next dimens
 ion: searching for epistasis -  Dr Emma Huang\, CSIRO\, Australia
DTSTART:20140610T150000Z
DTEND:20140610T160000Z
UID:TALK52883@talks.cam.ac.uk
CONTACT:Helen Mawdsley
DESCRIPTION:Multiparental populations have now been developed in several s
 pecies with the aim of facilitating the detection of novel gene-trait asso
 ciations. Most studies have focused on how best to capitalize on the incre
 ased diversity and precision of such populations through the detection of 
 QTL main effects. However\, the promise of multiparental populations\, in 
 large part\, lies in their creation of novel combinations of alleles which
  may lead to new insights regarding epistatic interactions. \n\nDetection 
 of epistatic effects is in part a computational issue - an exhaustive scan
  of pairwise comparisons may be several orders of magnitude more intensive
  than a one-dimensional scan. It is further complicated by statistical iss
 ues such as multicollinearity\, power\, and the most appropriate choice of
  model for individual founder effects. I will discuss different approaches
  to epistatic scans in multiparental populations and their feasibility at 
 a genomewide scale\, with examples from simulated data and application to 
 data from the Arabidopsis MAGIC population. In particular\, the ability to
  harness the power of parallel computing through Graphics Processing Units
  (GPUs) shows great promise for exhaustive interaction searches. 
LOCATION:Boardroom of The Sainsbury Laboratory (Bateman Street)
END:VEVENT
END:VCALENDAR