Algorithms for Molecular Kinetics

摘要 QUB是一种传统上用于分析单离子数据的计算工具 频道。它是一个能够处理刺激和数据记录的通用程序， 对数据进行模拟，并求解提取动力学的马尔可夫逆问题 生成数据的模型。在过去的15年里，它一直由美国国立卫生研究院和 NSF、凯克基金会和IBM也参与了这项研究。它被(免费)下载了数千次 每年一次，并在许多出版物中被引用。我们已经亲自培训了150多名 来自学术界、工业界和政府的学生每年都会使用该计划 当然了。近年来，该程序已被应用于数据的分析 分子马达、自旋电子学和小鼠的睡眠周期。它适用于任何系统 以马尔可夫动力学为特征，即状态模型。QUB是唯一可用于 研究人员，无论是商业的还是免费的，都可以在没有编程的情况下做到这一点。 QUB在过去的20年里一直在增长，现在包含大约30万条线路 用四种语言编写代码。许多已添加到解决方案中的自定义例程 直接的问题从来没有被记录在案，这些问题导致了一些深奥的代码 这并不容易被运输。此应用程序建议重写并记录QUB 源代码使用了Python来处理接口和C++的计算引擎。我们会 将新代码开源，以便用户可以进行修改。以减少 斜率的学习曲线，菜单会更有层次感，让最常见 只需尽可能少的鼠标/键盘敲击即可执行功能。我们会设置 UP QUB利用典型的松散耦合网络计算机 在实验室环境下，加快处理大数据集和多模型。 我们将创建向导来引导非专业用户完成模拟和分析，并创建 各种主题的教学DVD。这些视频可从以下地址下载 我们的维基网站。

Abstract QuB is a computational tool used traditionally to analyze the data from single ion channels. It is a general program capable of handling stimulus and data recording, the simulation of data, and solving the inverse Markov Problem of extracting the kinetic model that produced the data. It has been funded for the last 15 years by the NIH and also by NSF, the Keck Foundation, and IBM. It is downloaded (for free) thousands of times a year and is cited in many publications. We have personally trained more than 150 students from academia, industry and government in use of the program at an annual course. In recent years the program has been applied to the analysis of data from molecular motors, spintronics and the sleep cycles in mice. It applies to any system characterized by Markov kinetics, i.e. state models. QuB is the only program available to researchers, commercial or free, that can do this without programming. QuB has been growing for the last twenty years and now contains about 300,000 lines of code in four languages. Many of the custom routines that have been added to solve immediate problems were never documented and these have lead to some esoteric code that is not readily transported. This application proposes to rewrite and document QuB source code using Python to handle the interfaces and C++ the compute engines. We will make the new code open source so that users can make modifications. To decrease the slope of the learning curve, the menus will be more hierarchical so that the most common functions can be executed with the fewest possible mouse/keyboard strokes. We will set up QuB to utilize the loosely coupled networked computers found in the typical laboratory environment to speed up the processing of large data sets and multiple models. We will create Wizards to lead the lay user through simulation and analysis and create instructional DVDs on various topics. These videos will be available for download from our wiki site.