Comparisons between simulated and experimental free-running period shifts δτ in different constant light conditions and genetic backgrounds for plant models AP2012 and AP2012S.

For all light-mutant combinations considered, δτ was calculated as the difference between the mutant and wild-type periods: δτ = δτmut − δτwt. Hence, δτ δτ > 0 corresponds to a long-period mutant (denoted by a + sign). The study providing the experimental values used to calculate δτ is reported in the rightmost column in each case. In the above, Δ denotes knockout/knockdown, ‘OX’ denotes overexpression, ‘arr.’ denotes an arrhythmic oscillation, LL denotes constant light and DD denotes constant dark. Knockout/knockdown mutant behaviour was simulated by reducing the transcription rate of the target TF by 80%, with the exception of the ZTL mutant for which the protein production rate constant was decreased by 80% instead (there is no term for ZTL mRNA production in AP2012—see eqs. (S1.5) in S1 Text). Overexpression mutant behaviour was simulated by increasing the translation rate of the target TF two-fold. Simulated periods were calculated by using the MATLAB function findpeaks to obtain all the differences between two successive maxima in the circadian rhythm, and then averaging across all circadian gene components.