Quantitative Aspects of Single Molecule Detection

描述（由申请人提供）：允许在细胞环境中检测单分子的技术的出现，有望对细胞中发生的分子过程获得重要的新见解。虽然前景很好，但技术问题也很严重。这主要是因为与样品和检测系统中的重要噪声源相比，单个分子发出的信号较低。因此，单分子实验的任何定量分析都是具有挑战性的，需要精心开发工具。本提案的总体目标是增加单分子实验分析的可用方法。特别是，我们解决的基本问题使用的方法是新颖的单分子显微镜，但已成功地应用于其他工程学科。我们将从现代的角度分析单分子显微镜的性能限制。

DESCRIPTION (provided by applicant): The advent of technology that allows the detection of single molecules in a cellular environment holds the promise that significant new insights can be gained into molecular processes that occur in a cell. While the promise is great so are the technical problems. This is primarily because the signal emitted by a single molecule is low in comparison to the significant noise sources in the sample and in the detection system. Therefore any quantitative analysis of single molecule experiments is challenging and requires carefully developed tools. The overall goal of this proposal is to add to the available methods for the analysis of single molecule experiments. In particular, we address fundamental problems using methods that are novel to single molecule microscopy but have been successfully used in other engineering disciplines. We will analyze performance limits of single molecule microscopy from a modern point of view. Our specific aims are: 1.) To develop novel analytical tools to investigate the accuracy with which single molecules can be localized. 2.) To investigate existing algorithms and to develop novel algorithms for the localization of single molecules. 3.) To develop novel resolution criteria for single molecule detection and algorithms for the determination of the distance between two single molecules. 4.) To develop software modules for our existing Matlab based microscope image processing toolbox. We will use approaches based on the Fisher information matrix to obtain novel expressions for the accuracy with which a single molecule can be localized and for the accuracy with which the distance between two single molecules can be measured. Of particular importance is the determination to what extent experimental conditions such as cellular background and noise in the detection system deteriorate the localization and resolution accuracy. Existing algorithms will be evaluated to establish how close their performance is to the performance limit. Novel algorithms will be designed to meet the performance measures. A further criterion will be to assess the sensitivity of the algorithms to experimental artifacts such as outliers in the data. All proposed methods will be tested on immunological systems.