Biclustering genes and samples is important to extract knowledge from
gene expression measurements. Current methods are not generative
models allowing model selection and Bayesian framework or are additive
models not explaining mRNA effects and PCR amplification. We
introduce a generative, multiplicative model which assumes realistic
non-Gaussian heavy tail distributions.
Sprache der Kurzfassung:
Englisch
Erscheinungsmonat:
7
Erscheinungsjahr:
2010
Notiz zum Zitat:
Poster presentation at 18th Annual International Conference on Intelligent Systems for Molecular Biology (ISMB), Boston, MA USA, July 11 - July 13 2010