DNA microarray technology enables the simultaneous measurement of the transcript level of thousands of genes. Primary analysis can be done with basic statistical tools and cluster analysis, but effective and in depth analysis of the vast amount of transcription data requires integration with data from several heterogeneous data sources, such as upstream promoter sequences, genome-scale metabolic models, annotation databases and other experimental data. In this review, we discuss how experimental design, normalisation, heterogeneous data and mathematical modelling can enhance analysis of Saccharomyces cerevisiae whole genome transcription data. A special focus is on the quantitative aspects of normalisation and mathematical modelling approaches, since they are expected to play an increasing role in future DNA microarray analysis studies. Data analysis is exemplified with cluster analysis, and newly developed co-clustering methods, where the DNA microarray analysis is enhanced by integrating data from multiple, heterogeneous sources.