Statistical Mechanics of Gene Regulation

这项理论研究由材料研究和分子与细胞生物学部门联合资助。 基因组学和分子生物学技术的快速发展给分子生物学带来了许多令人兴奋的事情和机会。 大量的数据已经被收集在基因组序列、基因表达谱和蛋白质-蛋白质相互作用上。 涉及基因、蛋白质和细胞网络的系统级问题已经成为定量理解分子和细胞生物学的新挑战。 本研究将在基因调控的背景下研究系统级问题。 一个明确的模型将被用来描述基因表达的组合控制，并使用简单的通用相互作用的transciption因子与彼此和它们的DNA构建位点。该模型将用于研究如何将分子组分放在一起，以实现日益复杂的调控功能，以及这些调控系统如何相互耦合，形成表现出强大的基因表达模式的遗传网络。 该模型的一个关键特征是识别一个相互作用的子集，该子集可以通过简单地操纵调控区中的DNA序列来调整。 它们被称为“可编程”参数，可以由系统单独调整，以实现强加的功能要求。 该模型可以以相互作用的自旋系统的形式铸造，并且与神经网络中研究的“递归网络”有着惊人的相似之处。 在某些情况下，它也可以简化为更通用的布尔描述。 将进行分析和数值研究，以表征网络的特性，例如，系统可以实现的调节功能的复杂性，以及网络可以维持的稳定表达模式的能力。 特别是，一个随机的“监督学习规则”类似的监管序列的进化动力学将被研究，以搜索系统的可编程参数，最好地实现所需的功能反应。 因此，这种动力学也可以用来研究调控系统和基因网络的进化性。 这些调控和相互作用的基因系统的稳定性和复杂性的演变也将被研究。%这项理论研究由材料研究和分子与细胞生物学部门联合资助。 基因组学和分子生物学技术的快速发展给分子生物学带来了许多令人兴奋的事情和机会。 大量的数据已经被收集在基因组序列、基因表达谱和蛋白质-蛋白质相互作用上。 涉及基因、蛋白质和细胞网络的系统级问题已经成为定量理解分子和细胞生物学的新挑战。 本研究将在基因调控的背景下研究系统级问题。

This theoretical research is funded jointly by the Divisions of Materials Research and Molecular and Cellular Biology. Rapid advances in genomic and molecular technology have led to much excitement and opportunities in molecular biology. A hugh amount of data have already been collected on genomic sequences, gene expression profiles, and protein-protein interactions. System-level issues involving gene, protein, and cellular networks have emerged as the new challenges in quantitative understanding of molecular and cellular biology. This research will study system-level issues in the context of gene regulation. An explicit model will be used to describe the combinatoric control of gene expression, and using simple generic interactions of transciption factors with each other and with their DNA building sites. The model will be used to investigate how the molecular components may be put together to implement regulatory functions of increasing complexity, and how these regulatory systems can in turn be coupled to each other to form genetic networks exhibiting robust patterns of gene expression. A key feature of this model is the identification of a subset of interactions that can be tuned by simple manipulation of DNA sequences in the regulatory regions. They are referred to as "programmable" parameters and can be individually adjusted by the system in order to implement the imposed functional requirements. The model can be cast in the form of an interacting spin system, and shares striking similarities with the "recurrent networks" studied in neural networks. It can also be reduced to a more versatile Boolean description in some cases. Analytical and numerical studies will be done to characterize the properties of the networks, e.g., the complexity of the regulatory functions the system can implement, and the capacity of the stable expression patterns the network can maintain. In particular, a stochastic "supervised learning rule" resembling the evolutionary dynamics of regulatory sequences will be studied to search for the programmable parameters of the system that best achieve the desired functional responses. This dynamics can therefore be used also to study the evolvability of the regulatory control systems and the gene networks. The evolution of the stability and complexity of these regulatory and interactive gene systems will also be studied.%%%This theoretical research is funded jointly by the Divisions of Materials Research and Molecular and Cellular Biology. Rapid advances in genomic and molecular technology have led to much excitement and opportunities in molecular biology. A hugh amount of data have already been collected on genomic sequences, gene expression profiles, and protein-protein interactions. System-level issues involving gene, protein, and cellular networks have emerged as the new challenges in quantitative understanding of molecular and cellular biology. This research will study system-level issues in the context of gene regulation.***