About this Research Topic
Single-Molecule Localization Microscopy (SMLM) is a versatile and powerful super-resolution technique that allows researchers to study (biological) structures beyond the diffraction limit. One of the unique features of SMLM is that it provides not just a high-resolution image, but spatio-temporal, molecular coordinates for every detected emission event. As a result, SMLM is not only a method to unravel novel nanoscale structures, but is increasingly used as a quantitative approach for molecular clustering, tracking and stoichiometry. Consequently, analytical tools are essential components at all stages of a SMLM experiment. Getting from raw data to (biologically) meaningful results involves a number of processing steps that can roughly be divided into:
• detection, localization & filtering of raw data in order to build the multi-dimensional pool of data-points
• (meta-) analysis and visualization of said pool to generate interpretable high-resolution data
The SMLM community has developed a plethora of software packages to address the first part, with good reason: the quality of protein localization is, at best, only as good as the estimation method itself. However, the resulting multi-dimensional localization data enables unprecedented modes of post-processing analysis: while a classical optical microscope generates an image where each pixel's intensity is proportional to the amount of light hitting the detector, the intensity in the localization density map is proportional to the density of proteins, regardless (up to a point) of the intensity of the raw data. This lead to an ongoing paradigm change in the applied concepts for spatial analysis. Quantification moves from mere fluorescence intensity to molecular counting. Feature segmentation is now closely related to point-clustering. Denoising is further away from signal processing methods and closer to, again, clustering and/or spatial statistics.
Another important consequence of the high spatial resolution of SMLM is that it brings optical microscopy into the arena of electron microscopy (EM). Correlative methods, either between SMLM and EM, or also other super-resolution techniques, quickly become an enticing new way of studying life at the nanometer length-scale. Here, the promise lies in the complementary merging and analysis of cross-modality data, ranging from cross-modal registration at the nanometer scale and their implication for co-localization analyses, to multi-modal particle averaging, up to molecular counting in the context of nanometer ultrastructure.
Finally, while Deep Learning methods have already been applied to detection and localization, applications of these powerful universal approximators to the multi-dimensional localization set are very sparse and highly anticipated, as they could provide a new angle on the extraction of molecular information even beyond the scale that modern SMLM provides.
In-scope:
• Clustering
• High-resolution Image Segmentation
• Denoising (pre- and post- localization)
• Localization-based particle averaging methods, or related, to boost spatial resolution (like cryo EM)
• 3D Spatial Distribution Analysis (e.g. is a distribution uniform?)
• Visualization, Interactive Analysis tools
• Co-localization
• Correlative / Cross-modal analysis (EM+SMLM, Light-sheet+SMLM, STED+SMLM, ...)
• Deep Learning approaches to spatial data analysis
Out of scope:
• Novel peak detection algorithm
• Application of a well-established tool to solve a biological problem
• Novel imaging methods (hardware)
• detection, localization & filtering of raw data in order to build the multi-dimensional pool of data-points
• (meta-) analysis and visualization of said pool to generate interpretable high-resolution data
The SMLM community has developed a plethora of software packages to address the first part, with good reason: the quality of protein localization is, at best, only as good as the estimation method itself. However, the resulting multi-dimensional localization data enables unprecedented modes of post-processing analysis: while a classical optical microscope generates an image where each pixel's intensity is proportional to the amount of light hitting the detector, the intensity in the localization density map is proportional to the density of proteins, regardless (up to a point) of the intensity of the raw data. This lead to an ongoing paradigm change in the applied concepts for spatial analysis. Quantification moves from mere fluorescence intensity to molecular counting. Feature segmentation is now closely related to point-clustering. Denoising is further away from signal processing methods and closer to, again, clustering and/or spatial statistics.
Another important consequence of the high spatial resolution of SMLM is that it brings optical microscopy into the arena of electron microscopy (EM). Correlative methods, either between SMLM and EM, or also other super-resolution techniques, quickly become an enticing new way of studying life at the nanometer length-scale. Here, the promise lies in the complementary merging and analysis of cross-modality data, ranging from cross-modal registration at the nanometer scale and their implication for co-localization analyses, to multi-modal particle averaging, up to molecular counting in the context of nanometer ultrastructure.
Finally, while Deep Learning methods have already been applied to detection and localization, applications of these powerful universal approximators to the multi-dimensional localization set are very sparse and highly anticipated, as they could provide a new angle on the extraction of molecular information even beyond the scale that modern SMLM provides.
In-scope:
• Clustering
• High-resolution Image Segmentation
• Denoising (pre- and post- localization)
• Localization-based particle averaging methods, or related, to boost spatial resolution (like cryo EM)
• 3D Spatial Distribution Analysis (e.g. is a distribution uniform?)
• Visualization, Interactive Analysis tools
• Co-localization
• Correlative / Cross-modal analysis (EM+SMLM, Light-sheet+SMLM, STED+SMLM, ...)
• Deep Learning approaches to spatial data analysis
Out of scope:
• Novel peak detection algorithm
• Application of a well-established tool to solve a biological problem
• Novel imaging methods (hardware)
Keywords: Single-Molecule Localization Microscopy, SMLM, image segmentation, co-localization, image analysis, spatial data analysis
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
