Statistical inference for molecules: How many, when and where? (A07*)

在这个项目中，我们的目标是通过利用单个分子一次只能发射一个光子的事实，为STED等扫描显微镜绘制样品的空间分子和亮度分布。这将建立在高阶光子符合的严格统计卷积建模的基础上。基于这个模型，我们设想各种新的统计推断任务的发展，例如，在一个给定的点上的分子的数量，包括获得有关估计误差的信息的可能性。在实验方面，我们需要优化和调整显微镜的检测路径，并筛选新的染料。总之，我们的目标是开发一种新的自动化的统计分析工具，扫描显微镜。所获得的分子图谱及其统计置信限将为我们对有趣的生物现象提供新的见解。

In this project we aim to map the spatial molecular and brightness distribution of the sample for scanning microscopy such as STED, by exploiting the fact that a single molecule can emit only one photon at a time. This will build on a rigorous statistical convolution modeling of higher order photon coincidences. Based on this model we envision the development of various novel statistical inference tasks, e.g. on the number of molecules at a given spot, including the possibility to obtain information about the estimation error. On the experimental side we need to optimize and adapt the detection path of our microscopes and screen novel dyes. In summary, we aim to develop a novel automatized statistical analysis tool for scanning microscopy. The obtained molecular maps and their statistical confidence bounds will give us new insights into interesting biological phenomena.