Collaborative Research: Integrating Multiple Analyses to Understand Gene Regulatory Networks

基因调控网络在生命的几乎每一个过程中都扮演着至关重要的角色，因为它们负责接收环境刺激并将其转化为细胞内的活动。在这个项目中，PI将为基于计算机的建模开发新的策略，从而理解和预测复杂基因调控网络的行为。即将开发的计算机程序(算法)代表了一种将科学发现与计算创新相结合的新战略。虽然该算法正被用于探索真菌细胞对压力的反应的基因调控网络，但它应该会得出适用于其他生物系统的原理。模型将通过将模型预测与从实验中收集的数据进行比较来验证，然后这些数据将被用于进一步完善模型。该项目还包括通过采用跨学科、以问题为基础的学习方法来推进本科教育的举措，这种方法将由多个州的团队组成，也包括研究生。这个项目的目标是开发一种新的方法来建模基因调控网络。PI将检验这样的假设，即网络子集的初始实验特征将允许识别生物分子成分及其连接性，从而建立网络拓扑。然后，可以使用系统范围的时间进程测量来将该网络改进为能够做出准确系统预测的反应动力学模型。实验上易驯化的模式真菌Nidulans的细胞壁完整性信号通路将作为模型。这一途径通过激活修复机制来恢复细胞的完整性，从而对细胞壁损伤做出反应。由于蛋白激酶在介导细胞调节活动中起着关键作用，PI将专注于蛋白激酶的子集及其相关底物的发现，以初步组装一个基本的网络。随后，该系统将被实验扰动，以使用强大的转录、蛋白质组和磷蛋白质组平台来测量其动态响应。利用这些数据，PI将采取两步法来开发能够描述模型基因调控网络的动态行为的耦合常微分方程组的动态系统。首先，将测试和改进近似模型的集合方法。在第二步中，集合将作为种子种群，用于进化算法，以生成更精细和准确的模型。这个奖项由系统和合成生物学(生物科学局)和生物技术和生化生物工程(工程局)的项目共同资助。

Gene regulatory networks play a vital role in nearly every process of life as they are responsible for receiving environmental stimuli and translating these into activity inside the cell. In this project, the PIs will develop new strategies for computer-based modeling, and hence understanding and predicting, behavior of complex gene regulatory networks. The computer programs (algorithms) that will be developed represent a novel strategy integrating scientific discovery with computational innovation. Although the algorithm is being used to explore the gene regulatory networks of a fungal cell's response to stress it should lead to principles applicable to other biological systems. The model will be validated by comparing model predictions with data collected from experiments and this data will then be used to further refine the model. This project also includes initiatives to advance undergraduate education by employing an interdisciplinary problem-based learning approach that will be comprised of multi-state teams that also involve graduate students. The goal of this project is to develop a new approach for modeling gene regulatory networks. The PIs will test the hypothesis that initial experimental characterization of a network subset will permit identification of the biomolecular constituents and their connectivity, thus establishing network topology. System wide time-course measurements can then be used to refine this network into a reaction kinetic model capable of making accurate system predictions. The cell wall integrity signaling pathway of the experimentally tractable model fungus Aspergillus nidulans will serve as a model. This pathway responds to cell wall damage by activating repair mechanisms that restore cell integrity. Because protein kinases play a pivotal role in mediating cellular regulatory activities, the PIs will focus on a subset of kinases and the discovery of their associated substrates to initially assemble a rudimentary network. Subsequently the system will be experimentally perturbed for measuring its dynamic response using a robust transcriptomic, proteomic and phosphoproteomic platform. Using this data, the PIs will take a two-step approach to developing the dynamic system of coupled ordinary differential equations able to describe dynamic behavior of a model gene regulatory network. First, an ensemble approach of approximate models will be tested and refined. In the second step, the ensemble will act as the seed population for use in an evolutionary algorithm to generate a more refined and accurate model. The PIs will then validate the model by iterative comparisons of in silico predictions with experimental results.This award is co-funded by programs in Systems and Synthetic Biology (Directorate for Biological Sciences) and Biotechnology and Biochemical Bioengineering (Directorate for Engineering).