Figure 3:

“Sleep Analysis” outputs from PHASE. Shown here are plots of sleep, measured as minutes/user-defined bin, averaged across days for 12 individual flies (A), across all 32 flies (B) and across both days and flies (C) in the same manner as that shown for averaged activity, using the same time-series used for the analysis described in Fig. 2 (see Figs. 1A-iii and 2). Gray shaded regions in panels (A) through (C) indicate the scotophase of the light/dark cycle. The exported spreadsheets from the PHASE’s sleep analysis were used to compare total sleep, sleep bout numbers and duration across day and night in male CS flies. (D) As previously described, total minutes of sleep is higher during the night (V = 15; p = 6.38e-08). (E) The number of sleep bouts was significantly higher during the night (V = 153.5; p = 0.04). (F) Sleep bout duration was not significantly different between day and night (V = 241; p = 0.68). All statistical inferences in panels (D), (E) and (F) were drawn based on Wilcoxon’s matched pairs tests implemented in R. (G) Also shown are plots of sleep, measured as minutes/user-defined bin, averaged across flies for males (top) and females (bottom) for three consecutive LD cycles. To select a three-hour window around ZT06, three-hours was defined as the “Hours of Zeitgeber” and “Zeitgeber On” time was changed accordingly in the “Data Settings” input panel (see Fig. 1A–ii). This modified daytime window is shown in green, while the actual LD cycle during the experiment is indicated by the black and white bars underneath the plot. In this window we quantified total minutes of sleep for females and males. (H) As expected from previous work, total minutes of sleep in this three-hour window was significantly higher in males than in females as judged by a Mann-Whitney U test implemented in R (W = 133.5; p = 3.84e-07). For all panels: * < 0.05, ** < 0.01, *** < 0.001, and NS (Not Significant).