In mammals, circadian
periodicity has been described for gene expression in the hypothalamus and multiple peripheral tissues.
It is accepted that 1015 of all genes oscillate in a daily rhythm, regulated by an intrinsic molecular clock. Statistical analyses of
periodicity are limited by the small size of datasets and high levels of stochastic noise. Here, we propose a new approach applying digital signal processing algorithms separately to each group of genes oscillating in the same phase. Combined with the statistical tests for periodicity, this method identifies circadian baseline
oscillation in almost 100 of all expressed genes. Consequently, circadian oscillation in gene expression should be evaluated in any study related to biological pathways. Changes in gene expression caused by mutations or regulation of environmental factors (such as photic stimuli or feeding) should be considered in the context of changes in the amplitude and phase of genetic oscillations.