A proposal for direct-ordering gene expression data by self-organising maps

Microarray experiments are employed to simultaneously measure the expression level of thousands of genes. The possible applications of the available datasets include: inference of individual gene functions; identification of genes from specific tissues; analysis of the behaviour of gene expression levels under various environmental conditions and under different cell cycle stages; and characterisation of inappropriately transcribed genes and several genetic diseases. A fundamental step in the analysis of gene expression data is the detection of genes presenting similar expression patterns. Considering that microarray technology allows the inspection of a wide range of aspects related to the genome at once, and therefore thousands of genes may be involved in an experiment, new computational tools for data analysis and alternative visualisation strategies are crucial to understand and uncover information present in the data. In this work, the gene ordering problem is modelled as a shortest-path problem and we present an algorithm based on competitive learning for rearranging gene expression data in a linear order aiming to reveal trends in large amount of data. The effectiveness of the algorithm is attested by means of computational simulations performed on publicly accessible data sets. (C) 2004 Elsevier B.V. All rights reserved.