Flight Stability and Neural Feedback Control in Genetically Modified Flies

在获取大量行为数据的新技术和神经遗传学的进步的推动下，许多领域在理解动物行为方面的努力出现了新的势头。迫切需要开发复杂的量化模型，这些模型可以解释行为并做出可测试的预测。找出我们如何从基本的物理定律开始解释昆虫飞行的部分神经行为是这项拟议工作的中心主题。昆虫是第一个进化为会飞的人，飞是不会掉下来的。了解昆虫必须做什么才不会摔倒，为探索飞行物理学和神经反馈控制之间的联系提供了一条途径。这个项目将使用转基因果蝇来检验PI关于苍蝇转向肌肉对飞行稳定性的作用的猜想。除了实验，PI还将构建计算模型，可以预测观察到的完整果蝇和转基因果蝇的自由飞行行为。拟议的工作将包括三个核心部分：1)高分辨率自由飞行的3D跟踪，2)控制算法的计算研究，3)通过构建能够解释实验的有效模型来理解控制和肌肉活动。生物学领域以实验研究为主。这一关键贡献是开发了分析复杂生物系统的新方法。这项工作将提出新的问题，并对生物物理学、数学建模、神经科学、昆虫学、进化生物学和机器人学的研究产生直接影响。建议的工作为学生提供了一个从事跨学科工作的绝佳机会。关于自然界如何运作的发现将为与公众分享基础科学的价值提供一个令人信服的案例。拟议的工作基本上是将飞行物理学融入我们对控制神经电路的理解中。飞行回路的直接神经记录需要将昆虫拴住。一个挑战是将内部控制算法与自由飞行行为联系起来。另一个挑战是在完好无损的果蝇中区分不同的控制反馈方案。使用特定运动神经元沉默的转基因果蝇，可以让我们挑选出单个操纵肌的功能。对完整果蝇和转基因果蝇的飞行反射的分析将进一步导致控制方案的新模型，这些模型是飞行平衡的机械感觉反馈的基础。这项工作将首次将自由飞行的计算模型与神经遗传学的进步相结合，以破译苍蝇在自由飞行中的平衡反射。除了直接观察之外，这项工作还将为定量预测指导飞行行为提供一条新的途径。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

There is a new momentum in efforts across many fields to understand animal behavior, prompted by the new technology in acquiring large sets of behavioral data and the advancement in neural-genetics. Much needed is the development of sophisticated quantitive models that can explain behavior and can make testable predictions. Finding out how we can start from basic physical laws to explain parts of neural behavior of insect flight is the central theme of the proposed work. Insects were the first that evolved to fly, and to fly is not to fall. Understanding what insects must do so as not to fall provides a pathway to probe the connection between physics of flight and the neural feedback control. This project will test the PI's conjecture on the role of fly’s steering muscle on flight stability, using genetically modified fruit flies. In addition to experiments, the PI will construct computational models that can predict the observed free flight behavior of both intact and genetically modified flies. The proposed work will consist three core parts: 1) 3D tracking of free flight with high resolution, 2) computational studies of control algorithms, and 3) understanding the control and muscle activity by constructing effective models that can explain the experiments. The field of biology is dominated by experimental studies. This key contribution is the development of new methods for analyzing complex biological systems. The work will raise new questions and have a direct impact on research in physics of living organisms, mathematical modeling, neural science, entomology, evolutionary biology, and robotics. The proposed work offers an excellent opportunity for students to engage in interdisciplinary work. The findings on how nature works will provide a compelling case for sharing the value of basic science with the general public.The proposed work is fundamentally about integrating the physics of flight into our understanding of the neural circuitries for control. Direct neural recordings of flight circuitries require insects to be tethered. One challenge is to connect the internal control algorithms to the free flight behavior. Another challenge is to disentangle different control feedback schemes in intact flies. Using genetically modified flies with specified motor-neurons silenced allows us to single out the function of individual steering muscle. Analyses of the flight reflexes in intact and genetically modified flies will further lead to new models for control schemes that underlie the mechano-sensory feedback for flight equilibrium. This work will be the first to combine computational modeling of free flight with the progress in neuro-genetics to decipher a fly’s equilibrium reflex in free flight. The work will offer a new pathway for flight behavior guided by quantitive predictions, in addition to direct observations.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.