A molecular profile can be defined as a snapshot of a defined biological state. Such a profile can be represented by different types of biological identities (RNAs, proteins, metabolites, etc.) and it can be an explicit determination (exact numbers/concentrations) or a relative comparison of two states (e.g. young versus old, diseased versus healthy, and mutant versus wild type). High-throughput techniques, in particular DNA microarrays, have been widely used in recent years to generate genome-wide profiles associated with disease state or particular processes. A widespread problem, however, is the difficulty to distinguish cause from effect, to uncover which molecular changes reflected in a profile are the most important ones (i.e. drivers) and which are non-crucial (i.e. passengers).