一个详细的基因启动子动力学模型揭示了进入生产伸长是一个高度准时的过程
A detailed model of gene promoter dynamics reveals the entry into productive elongation to be a highly punctual process
摘要：
基因转录是一个涉及数千个反应的随机过程。第一组反应发生在基因启动子附近，被认为是随机噪声背景下最重要的反应。最常见的转录模型主要关注激活子/阻遏子对总转录速率的影响，并将基础转录过程近似为一步事件。根据这种有效的模型，mRNA拷贝分布的Fano因子总是大于（超泊松）或等于1（泊松），并且低于该极限（亚泊松）的唯一方法是通过负反馈。部分由于这种限制，转录的第一阶段负责mRNA拷贝数中的大部分随机噪声。然而，通过考虑建立和驱动基础转录机制的所有主要反应，从结合启动子的第一个蛋白质到转录复合物（TC）进入生产性延伸，表明转录的前两个阶段，即前起始复合物（PIC）形成和启动子近端暂停（PPP），是一个高度准时的过程。换句话说，这个过程的第一步和最后一步之间的时间分布很窄，这导致了已经进入生产性伸长的TC数量的亚泊松分布。事实上，通过使用2000个不同的参数集和4种不同的反应网络拓扑结构的Gillespie算法模拟PIC形成和PPP，结果表明，只有4.4%的情况下，Fano因子> 1，上限为1.7，而对于31%的情况下，Fano因子低于0.5，下限为0.19。这些结果对mRNA分布中观察到的大多数随机噪声总是起源于启动子的观点提出了质疑。
英文摘要：
Gene transcription is a stochastic process that involves thousands of reactions. The first set of these reactions, which happen near a gene promoter, are considered to be the most important in the context of stochastic noise. The most common models of transcription are primarily concerned with the effect of activators/repressors on the overall transcription rate and approximate the basal transcription processes as a one step event. According to such effective models, the Fano factor of mRNA copy distributions is always greater than (super-Poissonian) or equal to 1 (Poissonian), and the only way to go below this limit (sub-Poissonian) is via a negative feedback. It is partly due to this limit that the first stage of transcription is held responsible for most of the stochastic noise in mRNA copy numbers. However, by considering all major reactions that build and drive the basal transcription machinery, from the first protein that binds a promoter to the entrance of the transcription complex (TC) into productive elongation, it is shown that the first two stages of transcription, namely the pre-initiation complex (PIC) formation and the promoter proximal pausing (PPP), is a highly punctual process. In other words, the time between the first and the last step of this process is narrowly distributed, which gives rise to sub-Poissonian distributions for the number of TCs that have entered productive elongation. In fact, having simulated the PIC formation and the PPP via the Gillespie algorithm using 2000 distinct parameter sets and 4 different reaction network topologies, it is shown that only 4.4% give rise to a Fano factor that is > 1 with the upper bound of 1.7, while for 31% of cases the Fano factor is below 0.5, with 0.19 as the lower bound. These results cast doubt on the notion that most of the stochastic noise observed in mRNA distributions always originates at the promoter.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
