Dynamical phase diagram of an auto-regulating gene in fast switching conditions

Li Jia Chen, Ramon Grima

Research output: Contribution to journalArticlepeer-review

Abstract / Description of output

While the steady-state behaviour of stochastic gene expression with auto-regulation has been extensively studied, its time-dependent behaviour has received much less attention. Here, under the assumption of fast promoter switching, we derive and solve a reduced chemical master equation for an auto-regulatory gene circuit with translational bursting and cooperative protein-gene interactions. The analytical expression for the time-dependent probability distribution of protein numbers enables a fast exploration of large swaths of parameter space. For a unimodal initial distribution, we identify three distinct types of stochastic dynamics: (i) the protein distribution remains unimodal at all times; (ii) the protein distribution becomes bimodal at intermediate times and then reverts back to being unimodal at long times (transient bimodality) and (iii) the protein distribution switches to being bimodal at long times. For each of these, the deterministic model predicts either monostable or bistable behaviour and hence there exist six dynamical phases in total. We investigate the relationship of the six phases to the transcription rates, the protein binding and unbinding rates, the mean protein burst size, the degree of cooperativity, the relaxation time to the steady state, the protein mean and the type of feed backloop (positive or negative). We show that transient bimodality is a noise-induced phenomenon that occurs when protein expression is sufficiently bursty and we use theory to estimate the observation time window when it is manifest.
Original languageEnglish
Article number174110
JournalThe Journal of Chemical Physics
Issue number17
Early online date6 May 2020
Publication statusE-pub ahead of print - 6 May 2020


Dive into the research topics of 'Dynamical phase diagram of an auto-regulating gene in fast switching conditions'. Together they form a unique fingerprint.

Cite this