前馈型基因调控网络中噪声的传播机制研究

发布时间：2019-05-22 23:31

【摘要】：生物系统是由随机性和确定性规律支配着的,涨落和噪声己经渗透到了生物系统的各个层面,越来越多的证据显示,在很多生物过程中表达噪声起着重要作用。前馈型基因调控网络模体作为一种很典型的基因模块形式,近十年来其结构、功能和动力学特征越来越受到关注。大量实验表明:生命活动中的生化反应涉及到的分子数目都比较小,涨落和噪声非常明显,不同的网络模体对噪声的影响也不一样。因此,研究生化系统中噪声同前馈环结构的关系,追溯表达噪声的源头,理解它们在前馈调节路径中的角色,阐明噪声在前馈环中传播的机制与一般性原理,是当前定量生物学研究的重要内容之一。本文针对前馈型基因调控网络的动力学模型,利用随机动力学方法和计算机模拟技术,对前馈型基因调控网络的随机动力学性质以及噪声的传播机制进行了深入的分析,取得了以下研究成果:(1)以一致性前馈转录调控环路为研究对象,得到了前馈型基因调控网络随机噪声的传播特征和噪声分解原理。首先,基于一致性前馈型基因转录调控网络的动力学模型,将前馈型基因转录调控环路分解成主路和支路两部分,写出联合概率分布满足的化学主方程,利用线性噪声近似得到涨落-耗散公式,从而得出能描述随机变量统计性质的方差以及协方差的公式;接着,借助对数增益函数概念,对涨落-耗散公式进行规范化处理,解析推导了整个环路的总噪声、主路的噪声以及支路的噪声的理论表达式。最后,利用随机模拟对理论结果进行了验证。我们发现:与主路的噪声以及支路的噪声相比较,前馈环全模型可以有效地降低输出噪声水平;系统存在一个转换点,当系统在转换点以下时主路的噪声占主导,在转换点以上时支路的噪声占主导。我们以一个十分简单的信号转导模块为例子揭示了噪声如何产生及传递的机制,阐明了一致性前馈型基因调控网络的噪声分解原理。(2)进一步研究了各种前馈转录调控回路的输出噪声特性以及传播机制,包括“与门”、“或门”情况下的一致性以及非一致性前馈环。通过引入对数增益系数,运用线性噪声近似的方法导出了噪声分解的理论公式,利用随机模拟对理论结果进行了验证。根据噪声分解算法的理论和模拟结果分析前馈型基因网络模体噪声传递的非线性行为,得到了三个一般性的结论:首先,“与门”非一致性前馈环上游因子二级噪声传播为负值,即上游因子能够间接压制下游因子的噪声。其次,“或门”一致性前馈环上游因子的一级传播是非单调的,因此上游因子对于下游因子噪声的直接控制呈现非线性现象。最后,前馈环中支路为负调控时,下游因子的总噪声是单调上升的。我们归纳总结前馈调控网路的一般噪声特性和传播机制,为揭示噪声在功能和进化中的作用提供了一个初步的结果。
[Abstract]:Biological system is dominated by randomness and certainty. Fluctuations and noise have penetrated into all levels of biological system. More and more evidence shows that the expression of noise plays an important role in many biological processes. As a typical form of gene module, the structure, function and dynamics of feedforward gene regulatory network motif have attracted more and more attention in recent ten years. A large number of experiments show that the number of molecules involved in biochemical reactions in life activities is relatively small, the fluctuations and noise are very obvious, and the effects of different network motifs on noise are also different. Therefore, the relationship between noise and feedforward loop structure in biochemical system is studied, the source of noise is traced back, their roles in feedforward adjustment path are understood, and the mechanism and general principle of noise propagation in feedforward loop are expounded. It is one of the important contents of quantitative biology research at present. In this paper, aiming at the dynamic model of feedforward gene regulatory network, the random dynamic properties and noise propagation mechanism of feedforward gene regulatory network are deeply analyzed by using random dynamics method and computer simulation technology. The following results have been obtained: (1) taking the consistent feedforward transcriptional regulatory loop as the research object, the propagation characteristics and noise decomposition principle of random noise in feedforward gene regulatory networks are obtained. Firstly, based on the dynamic model of consistent feedforward gene transcriptional regulatory network, the feedforward gene transcriptional regulatory loop is decomposed into two parts: the main circuit and the branch, and the chemical master equation satisfied by the joint probability distribution is written. The fluctuation-dissipation formula is obtained by using linear noise approximation, and the variance and covariance which can describe the statistical properties of random variables are obtained. Then, with the help of the concept of logarithmic gain function, the fluctuation-dissipation formula is standardized, and the theoretical expressions of the total noise of the whole loop, the noise of the main circuit and the noise of the branch are derived analytically. Finally, the theoretical results are verified by random simulation. We find that compared with the noise of the main circuit and the branch, the full model of the feedforward ring can effectively reduce the output noise level. There is a conversion point in the system. When the system is below the conversion point, the noise of the main road is dominant, and the noise of the branch road is dominant when the conversion point is above the conversion point. We take a very simple signal transduction module as an example to reveal how noise is generated and transmitted. The noise decomposition principle of consistent feedforward gene regulatory networks is clarified. (2) the output noise characteristics and propagation mechanism of various feedforward transcriptional regulatory circuits, including "and gate", are further studied. Consistent and inconsistent feedforward rings in the case of "OR gate". By introducing logarithmic gain coefficient, the theoretical formula of noise decomposition is derived by using linear noise approximation method, and the theoretical results are verified by random simulation. According to the theory and simulation results of noise decomposition algorithm, the nonlinear behavior of modal noise transmission in feedforward gene network is analyzed, and three general conclusions are obtained: first, The second-order noise propagation of the upstream factor of the "gate" inconsistent feedforward loop is negative, that is to say, the upstream factor can indirectly suppress the noise of the downstream factor. Secondly, the first-order propagation of the upstream factor of the "OR gate" consistent feedforward ring is non-monotonous, so the upstream factor is nonlinear to the direct control of the downstream factor noise. Finally, when the branch in the feedforward ring is negatively regulated, the total noise of the downstream factor increases monotonously. We summarize the general noise characteristics and propagation mechanism of feedforward control networks, which provides a preliminary result for revealing the role of noise in function and evolution.
【学位授予单位】：华中师范大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：Q811.4

【相似文献】