Pattern Recognition in Mixed Spaces Pattern Recognition in Mixed Spaces. Clustering gene expression profiles into biologically sensible groups

A. CHERNOMORETZ

Punta del Este, Uruguay

Conferencia; Medynol 2008, XVI Conference on Non-equilibrium Statistical Mechanics and Nonlinear Physics; 2008

<!-- @page { size: 8.5in 11in; margin: 0.79in } P { margin-bottom: 0.08in } --> DNA microchips are devices which provide a comprehensive overview of the transcriptional state of a cell at a given time. They render surrogates of the mRNA concentration for the complete setof transcripts that are being exported from the nucleus into the cytoplasm to build the proteins required by the cell. A typical DNA microchip experiment produces a huge amount of information. Clustering techniques are usually employed to reveal common patterns of gene expression across different samples. The rationale behind this approach is that genes with similar profiles of activity will have related functions or will be regulated by common mechanisms. In this communication, making use of the Gene Ontology database – the de-facto standard for gene product annotations- we introduce a new clustering procedure which integrates transcriptional data and biological information about gene functions.