Fig. 5: Detection of circadian rhythms using backscatter measurements.

a–c Growth of Synechocystis wild type (WT) (a), Synechocystis ∆kaiA1B1C1 mutant strain (b), and Synechocystis ∆kaiA3B3C3 mutant strain (c) in LL after initial synchronization by dilution. Graphs display the average backscatter signal as a rolling average (solid line) at 730 nm with SD (shaded) from a representative experiment with 4–5 wells per strain. Polynomial regression (dashed line) was fitted to the data. d, e For each replicate within the experiments with wild type (d) and Synechocystis ∆kaiA1B1C1 (e), the raw backscatter signal was subtracted from the polynomial regression fit to remove the contribution of the overall growth of the cultures. The average difference was subtracted for normalization. Displayed is the average of this normalized backscatter (solid lines) with SD (shaded). Curves are smoothed. The dotted line indicates a simple harmonic cosine fit. f Same as (d, e), but the two experiments investigating backscatter in the ∆kaiA3B3C3 mutant are shown. g, h Same as (d, e) for further Synechocystis kai gene mutants without the cosine fits. All data displayed in (d–h) were collected from the same representative experiment (except for the additional Synechocystis ∆kaiA3B3C3 in f). Data from all three independent experiments are shown along with the cosine fit in Fig. S10 (except for Synechocystis ∆kaiA3B3C3). i To determine the effect on the first cycle after synchronization, the phase shift and relative amplitude of each Synechocystis mutant were calculated for the first peak (approx. between 21–29 h) for three independent experiments (Fig. S10a). The circles, triangles, and squares represent the first, second, and third experiments, respectively. Synechocystis ∆kaiA3B3C3 was not included in the third experiment. Marker scales are proportional to the number of replicate wells within one experiment (Exp. 1: n = 5 for all strains except ∆kaiA1B1C1 (n = 4), Exp. 2: all strains n = 4, Exp 3: all strains n = 5). Error boundaries were calculated using formulas (3) and (5) (see Methods section). The X-axis is discontinuous. j For strains that displayed dampened backscatter oscillations, we derived the length of the first period from the distance of the first trough to the first peak (see materials and methods) for the three experiments in Fig. S10a (Synechocystis ∆kaiA3B3C3 only two experiments). The backscattering period of the wild type was determined in the same way for comparison. Boxes are based on all replicate wells (indicated as n) from all three experiments (discriminated by the same symbol) and range from the first to the third quartile. Whiskers extend to the furthest data point within 1.5x the interquartile distance. The red line indicates the median and the dashed line indicates the mean. Results of the pairwise statistical tests are shown in Fig. S10c. Source data are provided as a Source Data file.