[Genomic and post-genomic approaches to cancer biomarker discovery - microarray standards at last]

The advent of microarray technology has enabled oncologists to investigate the expression of thousands of genes on the genetic basis of cancer. Gene-expression profiling studies have provided a molecular classification of cancer into clinically relevant subtypes, new tools to predict disease recurrence and response to different treatments. Analyzing these data can often be a quagmire, with endless discussion as to what the appropriate statistical analyses for any given experiment might be, while raising questions about the role of these techniques in clinical practice and patient management. For the analysis of data, many different methods and new computational algorithms have evolved. Here we describe state-of-the-art of gene-expression studies in clinical cancer, and consider both their current limitations and future promise.