Using Simulations to Explore Genetic Networks

该项目继续发展计算机方法，将基因相互作用的实验图转化为现实的、详细的、正式的数学模型，使生物学家能够探索基因是否以及如何在调控网络中相互作用，奠定蛋白质的条纹空间模式，这些模式预示着后来发展的分割身体计划。在世界各地的许多实验室进行了大约1200人年的实验研究，最终详细了解了所谓的片段极性基因如何在早期果蝇胚胎中相互交叉调节。这个项目将这些数据综合到一个系统级模型中，量化这些基因之间的相互作用，这些基因主要在果蝇中表现出来，但被认为是所有高等动物共有的，协同作用赋予了这个网络惊人的稳健性。也就是说，系统级模型预测，即使基因之间所有相互作用的强度发生巨大变化，神经网络也能继续正确地执行其任务。该项目需要对该案例研究进行细化，并创建另一种基因网络（神经源网络）的新模型。该项目将确定在第一个网络中发现的这种鲁棒性是否也出现在其他实验研究良好的网络中。它将使用计算机建模来确定这种健壮性是如何进化的。这项研究将包括在真正的果蝇胚胎上进行实验，以找出计算机模型预测的基因网络可能具有的鲁棒性是否确实存在于真正的胚胎中。活胚胎中的基因网络是否以及如何与遗传和环境变异相抗衡？胚胎真的会经历大量的“发育噪音”吗？如果是这样，发育机制是如何利用或使自己免受这种变异的影响的？这个跨学科的项目将尝试用实验发育生物学、计算机实现的激光扫描共聚焦显微数据的三维分析、详细的现实数学模型的实现、求解微分方程系统的计算机算法以及搜索/可视化其相关的高维参数空间来回答这些问题。

This project continues development of computer methods for translating experimentally charted maps of genetic interactions into realistic, detailed, formal mathematical models that allow biologists to explore whether and how genes, interacting in regulatory networks, lay down stripedspatial patterns of proteins which prefigure the segmented body plan that later develops. Approximately 1200 person-years of experimental research, in many labs worldwide, have culminated in a detailed understanding of how the so-called segment polarity genes cross regulate each other in early fruit fly embryos. This project synthesizes those data into a systems-level model that quantifies how the interactions between these genes, characterized principally in fruit flies, but thought to be shared by all higher animals, act synergistically to endow this network with astonishing robustness. That is, the systems level model predicts that the network can continue toperform its task correctly even when the strengths of all the interactions between its genes change greatly. The project entails refinement of that case study, and creation of a new model of another gene network (the neurogenic network). The project will ascertain whether this robustness,found in the first network so modeled occurs also in other experimentally well studied networks. It will use computer modeling to ascertain how such robustness could have evolved. The research will include experiments on real fruit fly embryos to find out whether the robustness, which the computermodel predicts the gene network could have, actually exists in real embryos. Do, and how do, gene networks in living embryos contend with genetic and environmental variation? Do embryos actually experience significant amounts of "developmental noise", and if so, how do developmental mechanisms either exploit or insulate themselves from such variation? This cross-disciplinaryproject will try to answer such questions using a blend of experimental developmental biology, computer-implemented 3-D analysis of laser scanning confocal micrographic data, implementation of detailed realistic mathematical models, computer algorithms for solving differential equationsystems and searching/visualizing their associated high-dimensional parameter spaces.