Related Experiment Video
Updated: Dec 31, 2025
Super-Resolution Imaging to Study Co-Localization of Proteins and Synaptic Markers in Primary Neurons
Published on: October 31, 2020
Bayesian analysis of data from segmented super-resolution images for quantifying protein clustering
Tina Košuta1, Marta Cullell-Dalmau2, Francesca Cella Zanacchi3
1The Quantitative BioImaging lab, Facultat de Ciències i Tecnologia, Universitat de Vic - Universitat Central de Catalunya, Vic, Spain. carlo.manzo@uvic.cat and University of Ljubljana, Ljubljana, Slovenia.
This study introduces a Bayesian method for analyzing super-resolution microscopy data, accurately quantifying molecular aggregates. The approach enhances precision, especially with limited data, revealing mixed populations of molecular structures in cells.
Area of Science:
- Biophysics
- Cell Biology
- Microscopy
Background:
- Super-resolution imaging advances visualization of nanoscale biological structures.
- Accurate quantitation of molecular organization and stoichiometry is crucial but challenging due to fluorescence stochasticity.
Purpose of the Study:
- To develop a robust Bayesian approach for precisely quantifying the relative abundance of molecular aggregates with varying stoichiometry from segmented images.
- To provide a reliable method for model selection in fitting fluorescence imaging data, improving parameter determination accuracy.
Main Methods:
- A Bayesian approach utilizing a nested sampling algorithm to fit distributions of molecular count proxies (e.g., localizations, intensity).
- Comparison of mixture models with increasing complexity to determine optimal component number and weights.
- Validation using in silico data and comparison with existing statistical methods.
Main Results:
- The Bayesian method accurately quantitates molecular aggregate stoichiometry, outperforming other statistical approaches.
- Demonstrated improved performance with small-statistics or incomplete datasets, enabling single-image analysis.
- Application to super-resolution imaging of dynein in HeLa cells confirmed mixed populations of single motors and higher-order structures.
Conclusions:
- The developed Bayesian method offers a robust tool for precise molecular quantitation in super-resolution microscopy.
- This approach significantly enhances the accuracy of parameter determination, particularly for challenging datasets.
- The findings support the presence of diverse dynein populations in cellular environments.
