Machine Learning In Computational Epigenetics

[Seminar cancelled]

Speaker:        Dr. Youngsook Lucy JUNG
                Center for Biomedical Informatics
                Harvard Medical School, USA

Title:          "Machine Learning In Computational Epigenetics"

Date:           Tuesday, 13 March 2012

Time:           3:00pm - 4:00pm

Venue:          Room 2463 (near lifts 25/26), HKUST

Abstract:

DNA is wrapped around nucleosomes which are composed of histone proteins.
DNA and various modified histones and non-histone proteins are called
chromatin, which serve to cell differentiation, gene regulation and other
key cellular processes including disease. These epigenetic modifications
of histones and non-histones create patterns that relate to various
functional elements in the genome such as gene transcription initiation,
elongation, enhancer and repression. Identifying patterns of combinations
of the epigenetic modifications that are common across the genome is a
useful way to understand genome organization. The grouped recurring
combinations of patterns are called chromatin states. Genome-wide
annotations of chromatin states provide better understanding of the
potential functions of underlying DNA sequences. In this talk, I will
provide several examples of how machine learning techniques such as HMM
and K-means algorithms can be used in this context, i.e., finding
combinatorial patterns of chromatin states.


********************
Biography:

Dr. Youngsook Lucy Jung is a research fellow at the Center for Biomedical
Informatics at Harvard Medical School, USA. Jung obtained her PhD from the
Health Science Technology of Massachusetts Institute of Technology, USA,
and received her BS and MS from Electrical Engineering and Computer
Science department of Korean Advanced Institute of Science and Technology
(KAIST), Korea. Her research interests are in understanding chromatin
structure and functions in various systems using machine learning and
computational analyses. She also works on developing computational methods
and tools for next-generation sequencing data.