Figure 4.

Circadian parameters for (A) Bmal1 and (B) Per2 oscillations following exposure to cancer cell-conditioned media. Shown are measures of strength of rhythmicity (top), period (middle), and amplitude (bottom). The measure of rhythmicity is the relative power in the 16- to 32-h band of the power spectral density. The period and amplitude are estimated by fitting each time series to a damped cosine curve. Data points are color-coded by dose; within each, data is separated by experiment (those to the left are from the first, in the middle are from the second, and to the right are from the third). The distribution of measures for each treatment is compared to that of the vehicle using a randomization test for difference in means (NS indicates “not significant,” ^*p < 0.05, ^**p < 0.01, and no indicator above a treatment indicates that there were too few data points for the test to have sufficient power). Time series that did not fit well to a damped cosine (GOF < 0.9) were excluded from period and amplitude evaluations. (Vehicle = FBS control, EMT6 = EMT6 cell-conditioned media, 4T1 = 4T1 cell-conditioned media).