Dataset supporting Speiser, Mueller et al Nature Methods 2021, Deep learning enables fast and dense single-molecule localization with high accuracy

Data supporting Speiser, Mueller et al Nature Methods 2021, Deep learning enables fast and dense single-molecule localization with high accuracy.

Software can be found at https://github.com/TuragaLab/DECODE

Abstract:

Single-molecule localization microscopy (SMLM) has had remarkable success in imaging cellular structures with nanometer resolution, but standard analysis algorithms necessitate the activation of only single isolated emitters which limits imaging speed and labeling density. Here, we overcome this major limitation using deep learning. We developed DECODE, a computational tool that can localize single emitters at high density in 3D with highest accuracy for a large range of imaging modalities and conditions. In a public software benchmark competition, it outperformed all other fitters on 12 out of 12 data-sets when comparing both detection accuracy and localization error, often by a substantial margin. DECODE allowed us to take fast dynamic live-cell SMLM data with reduced light exposure and to image microtubules at ultra-high labeling density. Packaged for simple installation and use, DECODE will enable many labs to reduce imaging times and increase localization density in SMLM.

Included datasets:

The datasets included here are used to extensively test and compare the performance of DECODE for a wide variety of different recording techniques and conditions.

01_fig2a_e_sim_density

Includes several sets of simulated data that can be used to evaluate the performance of localization algorithms.

Metrics_Density_X are simulations with different emitter densities (Fig. 2c,d,e) and Signal to Noise Ratios (SNR, Fig. 2c).

CRLB_SNR are simulations of isolated emitters used for comparisons with the Cramér–Rao bound (Fig. 2a)

RMSE_Sigma_corr are simulations of repeated samples of the camera model for identical emitter parameters to test the uncertainty estimates of DECODE (Fig. 2b).

All localizations and evaluations are collected in Fig2_collect_5.hdf5 and can be accessed as shown in Fig_draft_upload.ipynb.

02_fig4a_c_time.zip

The same sample of microtubules, labeled with anti-α tubulin primary and AF647 secondary antibodies, imaged with different UV activation intensities to resulting in different emitter densities per frame and acquisition times between 93 and 1120 s, while keeping the total number of localizations the same. Used to evaluate the performance of localization algorithms across emitter densities (Fig. 4a,b,c).

03_fig4d_live_golgi

Fast live-cell SMLM on the Golgi apparatus labeled withα-mannosidase II-mEos3.2. Used to evaluate the ability to process fast dynamic live-cell SMLM data with reduced light exposure. (Fig. 4d)

04_fig4e_live_ER

Fast live-cell SMLM on the endoplasmic reticulum labeled with calnexin-mEos3.2. (Fig. 4e)

05_fig4f_live

Fast live-cell SMLM on the nuclear pore complex proteinNup96-mMaple acquired in 3 seconds.(Fig. 4f)

06_fig4g_h_density.zip

Microtubules labeled with a high concentration of anti-α and anti-β tubulin primary and Alexa Fluor 647 secondary antibodies.Used to evaluate the performance of localization algorithms at ultra-high labeling densities. (Fig. 4g,h).

07_fig5_lls.zip

Localizations inferred with DECODE for a COS-7 cell imaged with LLS-PAINT microscopy (Legant at. al 2016).

08_SF3_SM3_SM4_fast_live_cell

Fast live-cell SMLM on the endoplasmic reticulum labeled with calnexin-mEos3.2 (Supplementary Fig. 3)