RI: Small: BCSP: The Whole Worm: A Brain-Body-Environment Model of Nematode Chemotaxis

这项工作的总体目标是构建、分析和实验测试无脊椎动物的大脑-身体-环境定向模型。即使是相对简单的动物，对它们的行为也表现出非凡的灵活性和健壮性，这对人工系统的设计非常有用。动物是如何达到这种困难的平衡的？要了解动物是如何作为整体运作的，需要建立完整动物的大脑-身体-环境模型。这反过来需要一个动物系统，为其提供详细的行为、生物力学和神经元信息。在这方面，关于线虫的行为、解剖学、遗传学、发育和神经连接的知识可能是所有动物中最完整的，目前在其神经系统的电生理学方面也取得了重大进展。成功还需要一种建模方法，这种方法可以在整个项目生命周期中与实验方法协同工作，并为解释系统的综合功能提供一个理论框架。为了论证全动物模型的可行性，本方案描述了一个构建和分析线虫大脑-身体-环境一体化模型的项目。线虫是一种盐趋化性形式，其方向的变化通过逐步连续的调整而朝向源头。这里描述的方法结合了连接组数据挖掘和一种新的分类方法，用于处理模型群体的参数优化和聚类技术，以及通过行为操作对代表性主题进行动态和信息论分析，神经元消融和光刺激，以及由一位实验同事进行的生物系统电生理研究。

The overall goal of this work is to construct, analyze and experimentally test a brain-body-environment model of orientation in an invertebrate. Even relatively simple animals exhibit a remarkable combination of flexibility and robustness to their behavior that would be extremely useful to the design of artificial systems. How do animals achieve this difficult balance? Understanding how animals operate as integrated wholes requires building brain-body-environment models of complete animals. This in turn requires an animal system for which detailed behavioral, biomechanical, and neuronal information is available. In this regard, knowledge of the behavior, anatomy, genetics, development and neural connectivity of the nematode worm C. elegans is perhaps the most complete of any animal, with significant progress now also being made on the electrophysiology of its nervous system. Success also requires a modeling methodology that can work cooperatively with experimental methods throughout the project lifecycle and that can provide a theoretical framework for interpreting the integrated functioning of the system. In order to demonstrate the feasibility of whole-animal modeling, this proposal describes a project to construct and analyze an integrated brain-body-environment model of C. elegans klinotaxis, a form of salt chemotaxis in which changes in direction are oriented towards the source through gradual continual adjustments. The approach described here combines connectome data-mining and a novel bracketing methodology, parameter optimization and clustering techniques for working with populations of models, and dynamical and information-theoretic analysis of representative motifs with behavioral manipulations, neuronal ablation and photo-stimulation, and electrophysiological studies on the biological system by an experimental colleague.