Hi, I'm looking to classify microarray data that comes in time-series measurements (3 time points per class). To my knowledge, there is no package in Bioconductor that supports classification of time-series data? If someone should know about such a package, or can hint me in the right direction of a non-BC package, that would be very much appreciated ! Thanks Piet (Belgium) University of Antwerp

Not in R, but this might be interesting for you Gaussian process clustering of multidimensional time series: http://stats.ma.ic.ac.uk/naheard/public_html/software/splinecluster/in dex.html Francois On Mon, 2007-02-19 at 17:29 +0100, Piet van Remortel wrote: > Hi, > > I'm looking to classify microarray data that comes in time-series > measurements (3 time points per class). > > To my knowledge, there is no package in Bioconductor that supports > classification of time-series data? > > If someone should know about such a package, or can hint me in the > right direction of a non-BC package, that would be very much > appreciated ! > > Thanks > > Piet (Belgium) > University of Antwerp > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor >

Dear Piet, you haven't even said whether you want to classify the genes, the biological samples, the time points, or ...? If it is classification of samples, have you tried a generic classification method (e.g. random forests, svm, Tibshirani's PAM) on features computed from your time course (e.g. average level, slope, similarity to a template profile). Best wishes Wolfgang ------------------------------------------------------------------ Wolfgang Huber EBI/EMBL Cambridge UK http://www.ebi.ac.uk/huber > > I'm looking to classify microarray data that comes in time-series > measurements (3 time points per class). > > To my knowledge, there is no package in Bioconductor that supports > classification of time-series data? > > If someone should know about such a package, or can hint me in the > right direction of a non-BC package, that would be very much > appreciated ! > > Thanks > > Piet (Belgium) > University of Antwerp >