Minimizing Structural Bias in Single-Molecule Super-Resolution Microscopy

Mazidi, Hesam; Nehorai, Arye; Lew, Matthew D.

doi:10.1101/293670

Cited by 4 publications

(10 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A rotationally fixed dipole corresponds to α = 0, while α = 90 • represents an isotropic molecule. Molecular blinking trajectories were simulated using a two-state Markov chain (7). We used a wavelength of 637 nm, NA = 1.4, and spatially uniform background.…”

Section: Extension To 3d Smlm a Natural Extension Of Wif To 3dmentioning

confidence: 99%

“…( 36), we use a first-order approximation of the exact PSF, as in Ref. (3). Recall that for any M ∈ M(R 2 ) we have:…”

Section: Parametersmentioning

confidence: 99%

“…Here, we use a simple implementation of our algorithm from Ref. (3) in Matlab (R2018a) on a desktop computer with 20 GB RAM and an Intel® Core™ i7-6700 CPU (3.40 GHz clock). We plot the execution time of computing WIF for various densities and two image sizes in Fig.…”

Section: E Quantifying the Stability Of Wifmentioning

confidence: 99%

“…Input localization a b ∆r [1] rG [1] rG [1] rG [1,1] rG [1,2] rG [1,3] rG [1,4] rG [1,5] rG [1,6] rG [1,7] rG [1,8] Perturbation rG [1] rG [1,1] rG [1,2] rG [1,3] rG [1,4] rG [1,5] rG [1,6] rG [1,7] rG [1,8] Perturbed source molecule rG [1,j] jth closest grid point to rG [1] ∆r [1,8] ∆r [1,7] ∆r [1,6] ∆r [1,1] ∆r [1,2] ζ [1,2] ∆r [1,3] ∆r…”

Section: Supplementary Note 9: Pupil Fittingmentioning

confidence: 99%

“…Existing metrics for assessing SMLM image quality can be categorized broadly into two classes: those that require knowledge of the ground-truth positions of fluorophores (e.g., Jaccard index and imaging DNA calibration "rulers") (6)(7)(8)(9), and those that operate directly on SMLM reconstructions alone, possibly incorporating information from other measurements (e.g., diffraction-limited imaging) (10)(11)(12). While these methods are able to provide summary or aggregate measures of performance, none of them directly measure the accuracy of individual localizations in an arbitrary SMLM dataset.…”

mentioning

confidence: 99%

See 4 more Smart Citations

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

Mazidi

Ding

Nehorai

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

The resolution and accuracy of various single-molecule localization microscopes (SMLMs) are routinely benchmarked using simulated data, calibration "rulers," or comparison to secondary imaging modalities. However, these methods cannot quantify the nanoscale accuracy of an arbitrary SMLM dataset. Here, we show that by computing measurement stability under a well-chosen perturbation with accurate knowledge of the imaging system, we can robustly quantify the confidence of individual localizations without ground-truth knowledge of the sample. We demonstrate our broadly-applicable method, termed Wasserstein-induced flux, for measuring the accuracy of various reconstruction algorithms directly on experimental 2D and 3D data of microtubules and amyloid fibrils. We further show that the estimated confidences can be used to evaluate the experimental mismatch of computational models, enhance the accuracy and resolution of reconstructed structures, and discover sample heterogeneity due to hidden molecular parameters. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 localization accuracy, statistical confidence, localization software, model mismatch, Wasserstein distance Single-molecule localization microscopy (SMLM) has be-1 come an important tool for resolving nanoscale structures 2 and answering fundamental questions in biology (1-4) and 3 materials science (5). SMLM uses repeated localizations of 4 blinking fluorescent molecules to reconstruct high-resolution 5 images of a target structure. In this way, quasi-static features 6 of the sample are estimated from noisy individual images cap-7 tured from a fluorescence microscope. These quantities, such 8 as fluorophore positions (i.e., a map of fluorophore density), 9blinking "on" times, emission wavelengths, and orientations, 10 influence the random blinking events that are captured within 11 an SMLM dataset. By using a mathematical model of the 12 microscope, SMLM reconstruction algorithms seek to estimate 13 the most likely set of fluorophore positions and brightnesses 14 (i.e., a super-resolution image) that is consistent with the 15 observed noisy images. 16A key question left unresolved by existing SMLM method-17 ologies is: How well do the SMLM data, i.e., the images of 18 blinking single molecules (SMs), support the super-resolved im-19 age produced by an algorithm? That is, what is our statistical 20 confidence in each localization? Intuitively, one's interpreta-21 tion of an SMLM reconstruction could dramatically change 22 by knowing how trustworthy each localization is.23Existing metrics for assessing SMLM image quality can 24 be categorized broadly into two classes: those that require 25 knowledge of the ground-truth positions of fluorophores (e.g., 26Jaccard index and imaging DNA calibration "rulers") (6-9), 27 and those that operate directly on SMLM reconstructions 28 alone, possibly incorporating information from other measure-29 ments (e.g., diffraction-limited imaging) (10-12). While these 30 methods are able to provide summary or aggregate measures 31 of performan...

show abstract

Section: Extension To 3d Smlm a Natural Extension Of Wif To 3dmentioning

confidence: 99%

“…( 36), we use a first-order approximation of the exact PSF, as in Ref. (3). Recall that for any M ∈ M(R 2 ) we have:…”

Section: Parametersmentioning

confidence: 99%

Section: E Quantifying the Stability Of Wifmentioning

confidence: 99%

Section: Supplementary Note 9: Pupil Fittingmentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

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

Mazidi

Ding

Nehorai

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

4polar-STORM polarized super-resolution imaging of actin filament organization in cells

et al. 2022

View full text Add to dashboard Cite

Single-molecule localization microscopy provides insights into the nanometer-scale spatial organization of proteins in cells, however it does not provide information on their conformation and orientation, which are key functional signatures. Detecting single molecules’ orientation in addition to their localization in cells is still a challenging task, in particular in dense cell samples. Here, we present a polarization-splitting scheme which combines Stochastic Optical Reconstruction Microscopy (STORM) with single molecule 2D orientation and wobbling measurements, without requiring a strong deformation of the imaged point spread function. This method called 4polar-STORM allows, thanks to a control of its detection numerical aperture, to determine both single molecules’ localization and orientation in 2D and to infer their 3D orientation. 4polar-STORM is compatible with relatively high densities of diffraction-limited spots in an image, and is thus ideally placed for the investigation of dense protein assemblies in cells. We demonstrate the potential of this method in dense actin filament organizations driving cell adhesion and motility.

show abstract

Resolving the 3D rotational and translational dynamics of single molecules using radially and azimuthally polarized fluorescence

Zhang

Zhou

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

We report a radially and azimuthally polarized (raPol) microscope for high detection and estimation performance in single-molecule orientation-localization microscopy (SMOLM). With 5000 photons detected from Nile red (NR) transiently bound within supported lipid bilayers (SLBs), raPol SMOLM achieves 2.1 nm localization precision, 1.4° orientation precision, and 0.15 sr precision in estimating rotational wobble. Within DPPC SLBs, SMOLM imaging reveals the existence of randomly oriented binding pockets that prevent NR from freely exploring all orientations. Treating the SLBs with cholesterol-loaded methylβ-cyclodextrin (MβCD-chol) causes NR’s orientational diffusion to be dramatically reduced, but curiously, NR’s median lateral displacements drastically increase from 20.8 nm to 75.5 nm (200 ms time lag). These jump diffusion events overwhelmingly originate from cholesterol-rich nanodomains within the SLB. These detailed measurements of single-molecule rotational and translational dynamics are made possible by raPol’s high measurement precision and are not detectable in standard SMLM.

show abstract

Minimizing Structural Bias in Single-Molecule Super-Resolution Microscopy

Cited by 4 publications

References 47 publications

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

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

4polar-STORM polarized super-resolution imaging of actin filament organization in cells

Resolving the 3D rotational and translational dynamics of single molecules using radially and azimuthally polarized fluorescence

Contact Info

Product

Resources

About