The arrival of the 21st century has seen a transition in the conduct of science. The 60 years of NIH-funded investigator-driven experimentation, necessary to break down the complex nature of disease into bite-sized pieces, now creates a platform to perform integrative science. A major step in this new direction was the sequencing of the human and other model organism genomes. This has spawned many new –Omics, transcriptomics, proteomics and metabolomics, each re-invention of older disciplines. However, this technological improvement has not been accompanied by changes in experimental design or in the development of stable statistical measures of the quality of the data acquired in –Omics experiments. In order to extract meaningful information from –Omics experiments, it is first essential to remove all aspects of bias. This requires involvement of statisticians with experience in laboratory science to determine the sources of variation (the order of obtaining animal tissues, distribution of stored samples, order of carrying out analyses, including labeling with Cy-dye or iTRAQ reagents) and to design experimental approaches that balance variation out across experimental groups. Omics approaches should also ensure that the number of experimental replicates is adequate for providing statistical significance. All experiments contain false discovery. In a two-group experiment with one variable, we accept that a 5% chance or lower that the two groups are the same is sufficient to reject the null hypothesis. In –Omics experiments, often with thousands of variables, this assumption is not true. Under the null hypothesis, the expected p-values are uniformally distributed between 0 and 1. Therefore, the false discovery rate (FDR) is predictable. If the number of true changes can be estimated and a particular FDR accepted, the statistician can determine the p-value for an experiment. With this insight, the number of replicates in an experiment can be proposed. In two independent studies carried out by CNGI investigators on changes in gene expression in the mammary gland between days 21 and 50 involving group sizes of 8 animals, microarray analysis determined there were 4982 genes with an FDR of <1% in experiment 1; in experiment 2, 3032 of these genes have a FDR <1%. A high correlation in fold changes was observed, indicating that meaningful interpretations could be made of these data. Pathway analysis revealed that the changes were concentrated in energetics and metabolism, suggesting that the onset of puberty in the mammary gland represents alterations in master switches under transcriptional control.