BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Machine Learning Methods for Uncovering cis-Regulatory Modules - D
 r Mark Craven\, University of Wisconsin
DTSTART:20061129T140000Z
DTEND:20061129T150000Z
UID:TALK5865@talks.cam.ac.uk
CONTACT:Danielle Stretch
DESCRIPTION:The process of transcription is often regulated by systems of 
 factors\nwhich bind in specific arrangements\, called cis-regulatory modul
 es\n(CRMs)\, in promoter regions.  I will describe a discriminative\,\nmac
 hine-learning approach we have developed for uncovering CRMs in the\npromo
 ter regions of genes that seem to be co-regulated.  Our approach\nsimultan
 eously learns models of binding-site motifs as well as the\nlogical struct
 ure and spatial preferences of these motifs in CRMs.\nOur results on yeast
  data sets show better predictive accuracy than a\ncurrent state-of-the-ar
 t approach on the same data sets.  Our results\non yeast\, fly\, and human
  data sets show that the inclusion of logical\nand spatial aspects improve
 s the predictive accuracy of our learned\nmodels.
LOCATION:MR5\, DAMTP
END:VEVENT
END:VCALENDAR