Spotlights
Direct observation of motor protein stepping in living cells using MINFLUX
We developed MINFLUX tracking in 2D and 3D to study the 16 nm steps of the motor protein kinesin while it walks on microtubules in living cells. This is a first step towards monitoring functional conformational changes of protein machines at high spatiotemporal resolution in living systems (Deguchi et al., Science, 2023 - free full PDF version).
Nanoscale architecture of clathrin-mediated endocytosis in yeast
We developed high-throughput super-resolution microscopy and measured the nanoscale distribution of 23 endocytic proteins from >100 000 snapshots of endocytic structures in budding yeast, providing new insights into structure, assembly process, and force generation of the endocytic machinery (Mund et al., Cell, 2018).
Quantitative analysis of super-resolution data
How can we analyze huge super-resolution data sets in a meaningful way? We developed LocMoFit, a tool that enables fitting of super-resolution microscopy data to an arbitrary geometric model. The fit extracts quantitative parameters of individual cellular structures, which can be used to investigate dynamic and heterogenous protein assemblies and to create average protein distribution maps (Wu et al., Nature Methods, 2023).
A dynamic 3D reconstruction of the clathrin coat during endocytosis
We used 3D single-molecule localization microscopy to measure the precise geometry of the clathrin coat at large numbers of endocytic sites. Through pseudo-temporal sorting, we determined the average trajectory of clathrin remodeling during endocytosis. We developed a new physical model that describes the measured shapes and dynamics and could represent a general mechanism for clathrin coat remodeling (Mund, Tschanz et al., Journal of Cell Biology, 2023).
Deep learning for single-molecule localization
We developed a software that uses simulator-based inference to localize fluorophores in 3D at high densities, increasing the speed of single-molecule localization microscopy by one order of magnitude. This software outperformed all other software in a public software benchmark on all modalities (Speiser, Müller et al., Nature Methods, 2021).
14:00 - 15:00
IMP Lecture Hall 