BIC: Cellular Inspired Garbage Collection: Ubiquitin System and Memory Management

垃圾收集是许多现代计算机编程语言的关键组件。垃圾回收将程序员从显式内存管理的负担中解放出来，并且对于面向对象编程范例中的函数数据对象的独立性是必要的。遗憾的是，传统的垃圾收集算法并不适合实时嵌入式系统中的并发操作。另一方面，细胞系统还必须能够识别和回收不再使用、损坏或不完整的蛋白质。蛋白质的选择性破坏和再循环是由细胞内的泛素系统完成的。要被选择性破坏的蛋白质首先通过一系列的酶反应识别并标记泛素。然后这些蛋白质在蛋白质小体中被降解，它们的组成氨基酸被循环用于构建新的蛋白质。我们建议使用混合系统来模拟泛素系统中的酶途径，以深入了解它们的计算特性。我们的建模和模拟工作将描述泛素连接酶和蛋白酶体如何相互作用，选择性地识别适当的蛋白质底物。我们在泛素系统上的结果将有助于激励计算机系统中用于内存管理的新的垃圾识别算法。这些算法将基于分类器的分层组合，这些分类器识别记忆中对象的特征特征。这项提议的教育目标与我们的研究议程密切相关。我们希望向学生和其他研究人员介绍将生物分子系统视为计算系统的重要性。泛素系统将在研讨会和未来的研讨会上为学生提供教学演示，并通过我们与富兰克林学院、该大学的AMP暑期项目和我们的年度女孩机器人研讨会的合作，为公众社区提供教学演示。其他教育活动将有助于为工程学、生物科学和医学界的研究人员提供信息并开启对话。

Garbage collection is a crucial component of many modern computer programming languages. Garbage collection frees programmers from the burden of explicit memory management, and is necessary for the independence of functional data objects in object-oriented programming paradigms. Unfortunately, traditional garbage collection algorithms are not ideally suited for concurrent operation in real-time embedded systems. On the other hand, cellular systems must also be able to identify and reclaim proteins that are no longer being used or are damaged or incomplete. The selective destruction and recycling of proteins is accomplished by the ubiquitin system in the cell. Proteins to be selectively destroyed are first identified and tagged with ubiquitin through a series of enzymatic reactions. These proteins are then degraded in the proteosome, and their constituent amino acids are recycled for the construction of new proteins.We propose using hybrid systems to model the enzymatic pathways in the ubiquitin system to gain insight into their computational properties. Our modeling and simulation efforts will describe how the ubiquitin conjugation ligases and the proteosome interact to selectively recognize the appropriate protein substrates. Our results on the ubiquitin system will help motivate new garbage recognition algorithms for memory management in computer systems. These algorithms will be based upon the hierarchical composition of classifiers that recognize characteristic features of the objects in memory.The educational goals of this proposal are closely tied to our research agenda. We hope to introduce students as well as other researchers to the importance of viewing biomolecular sytems as computational systems. The ubiquitin system will serve as a teaching illustration for students in seminars and future workshops, as well as for the public community through our collaboration with the Franklin Institute, the university's AMP summer program, and our annual Robotics for Girls workshop. Other educational activities will serve to inform and open dialogue between researchers in the engineering, biological sciences, and medical communities.