A dynamic defect Totally Asymmetric Simple Exclusion Process (ddTASEP) model of transcription with nucleosome induced pausing.

License: ddTASEP_Algorithm, ddTASEP with Static Defects

Nucleosomes are recognized as key regulators of transcription. However, the relationship between slow nucleosome unwrapping dynamics and bulk transcriptional properties has not been thoroughly explored. Here, an agent-based model that we call the dynamic defect Totally Asymmetric Simple Exclusion Process (ddTASEP) was constructed to investigate the effects of nucleosome-induced pausing on transcriptional dynamics. Pausing due to slow nucleosome dynamics induced RNAPII convoy formation, which would cooperatively prevent nucleosome rebinding leading to bursts of transcription. The mean first passage time (MFPT) and the variance of first passage time (VFPT) were analytically expressed in terms of the nucleosome rate constants, allowing for the direct quantification of the effects of nucleosome-induced pausing on pioneering polymerase dynamics. The mean first passage elongation rate γ(h_c,h_o ) is inversely proportional to the MFPT and can be considered to be a new axis of the ddTASEP phase diagram, orthogonal to the classical αβ-plane (where α and β are the initiation and termination rates). Subsequently, we showed that, for β=1, there is a novel jamming transition in the αγ-plane that separates the ddTASEP dynamics into initiation-limited and nucleosome pausing-limited regions. We propose analytical estimates for the RNAPII density ρ, average elongation rate v, and transcription flux J in these regions that converge to the classical TASEP behavior in the limit γ→1 and verified them numerically. Finally, we demonstrate that the intra-burst RNAPII waiting times t_in follow the time-headway distribution of a max flux limit TASEP, that the average inter-burst interval (t_IBI ) correlates with the index of dispersion D_e and is inversely proportional to γ. In the limit γ→0, the average burst size reaches a maximum set by the closing rate h_c. Last, for cases with α≪1, the burst sizes are geometrically distributed, allowing large bursts even while the average burst size (N_B ) is small.