Automated analysis of multivariate nonlinear gene relations based on cDNA microarray expression data

A cDNA microarray is a complex biochemical-optical system whose purpose is the simultaneous measurement of gene expression for thousands of genes. This paper describes a general statistical environment for finding associations among gene expression patterns, and between genes and external conditions, via the coefficient of determination. This coefficient measures the degree to which the transcriptional levels of an observed gene set can be used to improve the prediction of the transcriptional state of a target gene relative to the best possible prediction in the absence of observations. Various aspects of the method are discussed: prediction quantification, design of predictors given small numbers of replicated microarrays, and constrained prediction using ternary perceptrons. A main focus is the supporting software and its facilities for data analysis and visualization.