Physics at Virginia

"Application of Machine Learning Methods to Genome-Wide Maps of Histone Methylations"

Stefan Bekiranov , UVA Medical School
[Host: Eugene Kolomeisky]
The physical length of one copy of the human genome is a little over 1 meter. It is packaged into a nucleus, which is on the order of micrometers in diameter. This is achieved by wrapping the DNA around histones. In the last decade, many breakthroughs have lead to the understanding that these histones control subsets of genes that are turned on or off depending on chemical modifications on their tails. They accomplish this by controlling the accessibility of proteins—responsible for turning genes on—to DNA. This accessibility can be characterized by two states: open and closed. Remarkably, over 60 different locations on these tails are subject to at least one of eight types of chemical modifications. Recently, it has been shown that many of these modifications work together to robustly turn genes on or off; however, we are at the beginning of uncovering this complex control network. To shed light on this network, we apply computational methods, which identify statistically significant combinations, to genome wide maps of histone modifications. We indeed find that crosstalk among these modifications is extensive and predict novel combinations, which strongly synergize in our models, for further biochemical study.
Friday, April 29, 2011
4:00 PM
Physics Building, Room 204
Note special time.
Note special room.

 Slideshow (PDF)
 Add to your calendar

To add a speaker, send an email to phys-speakers@Virginia.EDU. Please include the seminar type (e.g. Colloquia), date, name of the speaker, title of talk, and an abstract (if available).