scholarly journals Precise measurement of nanoscopic septin ring structures in deep learning-assisted quantitative superresolution microscopy

2021 ◽  
Author(s):  
Amin Zehtabian ◽  
Paul Markus Müller ◽  
Maximilian Goisser ◽  
Leon Obendorf ◽  
Lea Jänisch ◽  
...  
Keyword(s):  
Deep Learning ◽  
Single Molecule ◽  
Dynamic Range ◽  
Simulated Data ◽  
Experimental System ◽  
Superresolution Microscopy ◽  
Septin Ring ◽  
Macromolecular Assemblies ◽  
Ring Structures ◽  
Quantitative Changes

The combination of image analysis and fluorescence superresolution microscopy methods allows for unprecedented insight into the organization of macromolecular assemblies in cells. Advances in deep learning-based object recognition enables the automated processing of large amounts of data, resulting in high accuracy through averaging. However, while the analysis of highly symmetric structures of constant size allows for a resolution approaching the dimensions of structural biology, deep learning methods are prone to different forms of bias. A biased recognition of structures may prohibit the development of readouts for processes that involve significant changes in size or shape of amorphous macromolecular complexes. What is required to overcome this problem is a detailed investigation of potential sources of bias and the rigorous testing of trained models using real or simulated data covering a wide dynamic range of possible results. Here we combine single molecule localization-based superresolution microscopy of septin ring structures with the training of several different deep learning models for a quantitative investigation of bias resulting from different training approaches and finally quantitative changes in septin ring structures. We find that trade-off exists between measurement accuracy and the dynamic range of recognized phenotypes. Using our trained models, we furthermore find that septin ring size can be explained by the number of subunits they are assembled from alone. Our work provides a new experimental system for the investigation of septin polymerization.

scholarly journals New Results on Radioactive Mixture Identification and Relative Count Contribution Estimation

Sensors ◽  
2021 ◽  
Vol 21 (12) ◽  
pp. 4155
Author(s):  
Bulent Ayhan ◽  
Chiman Kwan
Keyword(s):  
Deep Learning ◽  
Noise Source ◽  
Simulated Data ◽  
Nuclear Material ◽  
Machine Learning Algorithms ◽  
Detector Response ◽  
Nuclear Materials ◽  
Ratio Estimation ◽  
Analysis Software ◽  
Detector Distance

Detecting nuclear materials in mixtures is challenging due to low concentration, environmental factors, sensor noise, source-detector distance variations, and others. This paper presents new results on nuclear material identification and relative count contribution (also known as mixing ratio) estimation for mixtures of materials in which there are multiple isotopes present. Conventional and deep-learning-based machine learning algorithms were compared. Realistic simulated data using Gamma Detector Response and Analysis Software (GADRAS) were used in our comparative studies. It was observed that a deep learning approach is highly promising.

scholarly journals Understanding Complex Single Molecule Emission Patterns with Deep Learning

2019 ◽  
Vol 116 (3) ◽  
pp. 281a
Author(s):  
Peiyi Zhang ◽  
Sheng Liu ◽  
Abhishek Chaurasia ◽  
Donghan Ma ◽  
Michael J. Mlodzianoski ◽  
...  
Keyword(s):  
Deep Learning ◽  
Single Molecule

scholarly journals Extraction of accurate cytoskeletal actin velocity distributions from noisy measurements

2020 ◽  
Author(s):  
Cayla M. Miller ◽  
Elgin Korkmazhan ◽  
Alexander R. Dunn
Keyword(s):  
Single Molecule ◽  
Actin Filaments ◽  
Dynamical Behavior ◽  
Jump Process ◽  
Pairwise Distance ◽  
Actin Dynamics ◽  
Cell Cortex ◽  
Superresolution Microscopy ◽  
Distance Distributions ◽  
Primary Fibroblasts

Dynamic remodeling of the actin cytoskeleton allows cells to migrate, change shape, and exert mechanical forces on their surroundings. How the complex dynamical behavior of the cytoskeleton arises from the interactions of its molecular components remains incompletely understood. Tracking the movement of individual actin filaments in living cells can in principle provide a powerful means of addressing this question. However, single-molecule fluorescence imaging measurements that could provide this information are limited by low signal-to-noise ratios, with the result that the localization errors for individual fluorophore fiducials attached to filamentous (F)-actin are comparable to the distances traveled by actin filaments between measurements. In this study we tracked the movement F-actin labeled with single-molecule densities of the fluorogenic label SiR-actin in primary fibroblasts and endothelial cells. We then used a Bayesian statistical approach to estimate true, underlying actin filament velocity distributions from the tracks of individual actin-associated fluorophores along with quantified localization uncertainties. This analysis approach is broadly applicable to inferring statistical pairwise distance distributions arising from noisy point localization measurements such as occur in superresolution microscopy. We found that F-actin velocity distributions were better described by a statistical jump process, in which filaments exist in mechanical equilibria punctuated by abrupt, jump-like movements, than by models incorporating combinations of diffusive motion and drift. A model with exponentially distributed time- and length-scales for filament jumps recapitulated F-actin velocity distributions measured for the cell cortex, integrin-based adhesions, and actin stress fibers, indicating that a common physical model can potentially describe F-actin dynamics in a variety of cellular contexts.

scholarly journals Quantifying accuracy and heterogeneity in single-molecule super-resolution microscopy

2019 ◽  
Author(s):  
Hesam Mazidi ◽  
Tianben Ding ◽  
Arye Nehorai ◽  
Matthew D. Lew
Keyword(s):  
Single Molecule ◽  
Computational Models ◽  
Amyloid Fibrils ◽  
Imaging System ◽  
Simulated Data ◽  
Super Resolution ◽  
Ground Truth ◽  
Localization Algorithm ◽  
Reconstruction Algorithms ◽  
Imaging Data

The resolution and accuracy of single-molecule localization micro-scopes (SMLMs) are routinely benchmarked using simulated data, calibration “rulers,” or comparisons to secondary imaging modalities. However, these methods cannot quantify the nanoscale accuracy of an arbitrary SMLM dataset. Here, we show that by computing localization stability under a well-chosen perturbation with accurate knowledge of the imaging system, we can robustly measure the confidence of individual localizations without ground-truth knowledge of the sample. We demonstrate that our method, termed Wasserstein-induced flux (WIF), measures the accuracy of various reconstruction algorithms directly on experimental 2D and 3D data of microtubules and amyloid fibrils. We further show that WIF confidences can be used to evaluate the mismatch between computational models and imaging data, enhance the accuracy and resolution of recon-structed structures, and discover hidden molecular heterogeneities. As a computational methodology, WIF is broadly applicable to any SMLM dataset, imaging system, and localization algorithm.

scholarly journals Publisher Correction: Deep learning enables fast and dense single-molecule localization with high accuracy

2021 ◽  
Author(s):  
Artur Speiser ◽  
Lucas-Raphael Müller ◽  
Philipp Hoess ◽  
Ulf Matti ◽  
Christopher J. Obara ◽  
...  
Keyword(s):  
Deep Learning ◽  
Single Molecule ◽  
High Accuracy

Fluorescence principles and measurement

2000 ◽  
Author(s):  
Arthur G. Szabo
Keyword(s):  
Fluorescence Spectroscopy ◽  
Single Molecule ◽  
Dynamic Range ◽  
Spectroscopic Method ◽  
Energy Difference ◽  
Visible Spectral Range ◽  
Analytical Determination ◽  
Whole Cells ◽  
Genomic Characterization ◽  
High Level

Fluorescence spectrometry is the most extensively used optical spectroscopic method in analytical measurement and scientific investigation. During the past five years more than 60000 scientific articles have been published in which fluorescence spectroscopy has been used. The large number of applications ranges from the analytical determination of trace metals in the environment to pH measurements in whole cells under physiological conditions. In the scientific research laboratory, fluorescence spectroscopy is being used or applied to study the fundamental physical processes of molecules; structure-function relationships and interactions of biomolecules such as proteins and nucleic acids; structures and activity within whole cells using such instrumentation as confocal microscopy; and DNA sequencing in genomic characterization. In analytical applications the use of fluorescence is dominant in clinical laboratories where fluorescence immunoassays have largely replaced radioimmunoassay techniques. There are two main reasons for this extensive use of fluorescence spectroscopy. Foremost is the high level of sensitivity and wide dynamic range that can be achieved. There are a large number of laboratories that have reported single molecule detection. Secondly, the instrumentation required is convenient and for most purposes can be purchased at a modest cost. While improvements and advances continue to be reported fluorescence instrumentation has reached a high level of maturity. A review of the physical principles of the fluorescence phenomenon permits one to understand the origins of the information content that fluorescence measurements can provide. A molecule absorbs electromagnetic radiation through a quantum mechanical process where the molecule is transformed from a ‘ground’ state to an ‘excited’ state. The energy of the absorbed photon of light corresponds to the energy difference between these two states. In the case of light in the ultraviolet and visible spectral range of 200 nm to 800 nm that corresponds to energies of 143 to 35.8 kcal mol-1. The absorption of light results in an electronic transition in the atom or molecule. In atoms this involves the promotion of an electron from an outer shell orbital to an empty orbital of higher energy.

scholarly journals Automatic classification and segmentation of single-molecule fluorescence time traces with deep learning

2020 ◽  
Vol 11 (1) ◽  
Author(s):  
Jieming Li ◽  
Leyou Zhang ◽  
Alexander Johnson-Buck ◽  
Nils G. Walter
Keyword(s):  
Deep Learning ◽  
Single Molecule ◽  
Resonance Energy Transfer ◽  
Resonance Energy ◽  
Markov Modeling ◽  
Single Molecule Fluorescence ◽  
Hidden Markov Modeling ◽  
Improved Performance ◽  
Automated Screening ◽  
Expectation Bias

AbstractTraces from single-molecule fluorescence microscopy (SMFM) experiments exhibit photophysical artifacts that typically necessitate human expert screening, which is time-consuming and introduces potential for user-dependent expectation bias. Here, we use deep learning to develop a rapid, automatic SMFM trace selector, termed AutoSiM, that improves the sensitivity and specificity of an assay for a DNA point mutation based on single-molecule recognition through equilibrium Poisson sampling (SiMREPS). The improved performance of AutoSiM is based on accepting both more true positives and fewer false positives than the conventional approach of hidden Markov modeling (HMM) followed by hard thresholding. As a second application, the selector is used for automated screening of single-molecule Förster resonance energy transfer (smFRET) data to identify high-quality traces for further analysis, and achieves ~90% concordance with manual selection while requiring less processing time. Finally, we show that AutoSiM can be adapted readily to novel datasets, requiring only modest Transfer Learning.

scholarly journals MIBI-TOF: A multiplexed imaging platform relates cellular phenotypes and tissue structure

2019 ◽  
Vol 5 (10) ◽  
pp. eaax5851 ◽  
Author(s):  
Leeat Keren ◽  
Marc Bosse ◽  
Steve Thompson ◽  
Tyler Risom ◽  
Kausalia Vijayaragavan ◽  
...  
Keyword(s):  
Single Molecule ◽  
Spatial Information ◽  
Dynamic Range ◽  
Ion Beam ◽  
Time Of Flight ◽  
Tissue Structure ◽  
Regional Variability ◽  
Flight Mass Spectrometry ◽  
Cellular Phenotypes ◽  
And Function

Understanding tissue structure and function requires tools that quantify the expression of multiple proteins while preserving spatial information. Here, we describe MIBI-TOF (multiplexed ion beam imaging by time of flight), an instrument that uses bright ion sources and orthogonal time-of-flight mass spectrometry to image metal-tagged antibodies at subcellular resolution in clinical tissue sections. We demonstrate quantitative, full periodic table coverage across a five-log dynamic range, imaging 36 labeled antibodies simultaneously with histochemical stains and endogenous elements. We image fields of view up to 800 μm × 800 μm at resolutions down to 260 nm with sensitivities approaching single-molecule detection. We leverage these properties to interrogate intrapatient heterogeneity in tumor organization in triple-negative breast cancer, revealing regional variability in tumor cell phenotypes in contrast to a structured immune response. Given its versatility and sample back-compatibility, MIBI-TOF is positioned to leverage existing annotated, archival tissue cohorts to explore emerging questions in cancer, immunology, and neurobiology.

scholarly journals Single-Molecule Imaging and Labelling with DNA-Based Superresolution Microscopy

2020 ◽  
Vol 118 (3) ◽  
pp. 146a
Author(s):  
Mingjie Dai ◽  
Ninning Liu ◽  
Sinem K. Saka ◽  
Peng Yin
Keyword(s):  
Single Molecule ◽  
Single Molecule Imaging ◽  
Superresolution Microscopy
Sign in / Sign up

Export Citation Format

Share Document