Genome informatics for data-driven biology.docVIP

/2002/3/4/reports/4010.1 Meeting report Genome informatics for data-driven biology Kenta Nakai* and Jean-Philippe Vert? Addresses: *Human Genome Center, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minatoku, Tokyo 108-8639, Japan.?Bioinformatics Center, Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan. Correspondence: Kenta Nakai. E-mail: knakai@ims.u-tokyo.ac.jp Published: 27 March 2002 GenomeBiology 2002, 3(4):reports4010.1–4010.3 The electronic version of this article is the complete one and can be found online at /2002/3/4/reports/4010 ? BioMed Central Ltd (Print ISSN 1465-6906; Online ISSN 1465-6914) where gene(X) represents the expression value of gene X. A report on the 12th International Conference on Genome This rule turned out to hold in 75% of cancer cells but in 0% Informatics, Tokyo, Japan, 17-19 December 2001. in normal tissues. Essentially, he developed a novel method (?entropy-oriented discretization method?) to automatically find such rules, including the determination of the threshold Genome informatics as a field encompasses the various methods and algorithms for analyzing and extracting bio- logically relevant information from the rapidly growing bio- logical - and especially genome sequence - databases. This leads to a new data-driven research paradigm for post- values. His method showed better performance than previ- ously reported methods of microarray analysis; furthermore, the method seems to be applicable to many other problems, including the prediction of protein subcellular localization. genomic biomedical research, which Charles (Chip) Lawrence (Wadsworth Center, Albany, USA), speaking at the meeting, claimed is replacing the traditional hypothesis- Comparing gene networks Several relationships can be formulated between genes in a given organism, such as those based on structural or func- tional similarity