Giosuè Lo Bosco, University of Palermo, Italy

Machine Learning Methods for Epigenomic Analyis
Epigenetics concerns the study of heritable factors on gene expression that are not directly coded in DNA sequences. Recent studies has shown that one of the major epigenetics factor is Nucleosomes spacial organization, where a nuclesome can be considered as the basic unit of eukaryotic chromatin and physically consists of about 150 bp of DNA wrapped around an histone proteins core.
To measure nucleosome positions on a genomic scale both theoretical and experimental approaches have been recently developed. Such methodologies are mainly based on probabilistic models able to infer nucleosome positions basing purely on genomic sequence information or on genomic-scale hybridization data achieved by a tiled microarray approach. In this tutorial recent developed machine learning algorithms for nucleosome positioning will be analyzed and discussed, giving also emphases to their comparison in terms of computational complexity and classification accuracy.

References

  1. Segal E., Widom J., What controls nucleosome positions? Trends Genet. 25:8 (2009), 335-343.

  2. G.-C. Yuan, Y.J. Liu, M.F. Dion, M.D. Slack, L.F. Wu, S.J. Altschuler, O.J. Rando, Genome-scale identification of nucleosome positions in S. cerevisiae, Science 309 (2005), 626-630.

  3. V. Di Gesù, G. Lo Bosco, L. Pinello, G.C. Yuan, D.F.V. Corona,"A Multi-Layer Method to Study Genome-Scale Positions of Nucleosomes", Genomics 93:2 (2009), 140-145.

  4. G-C Yuan , J.S. Liu, Genomic Sequence Is Highly Predictive of Local Nucleosome Depletion, PLoS Comput Biol 4:1 (2008) , 164-174.