←
Zach Marin22 posts
Zach Marin avatar

Zach Marin

@zacsimile

hardware x software in microscopy | views are graphical projections | postdoc in Ries Lab @MaxPerutzLabs | @zacsimile.bsky.social | 🇺🇸 in 🇦🇹 | he/him

◷ Joined January 2017▧ 6 Photos and videos
Jonas Ries avatar
⟳ Zach Marin retweeted
Jonas Ries@JonasRies·Jul 28

Long time in the making and finally out. Great work by @haydnspass and @hitshah95 who wrote the code, Maxime and Nestor for the beautiful samples and @zacsimile for helping to wrap this up. Try it out!

[0/1] Excited to share our new preprint DECODE-Plex! After many people efforts, we developed a deep-learning framework for multi-channel SMLM, enabling simultaneous localization and color assignment of densely packed emitters across multiple channels. biorxiv.org/content/10.64898…

Post attachmentPost attachment
Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

These and many other simulations are covered in the paper. Example notebooks for each simulation are provided in the code base, and on Google Colab (see github.com/ries-lab/SimuFLUX…). We hope the community will find this useful and contribute their own examples using this package.

Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

For single molecule tracking, we optimized laser power, pattern size, photon limit and pattern dwell time to increase the measurable diffusion coefficient by 70% over that captured by the standard Abberior tracking sequence.

Post attachment
Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

We simulated DNA-PAINT imaging using an Abberior sequence file, and showed that quenched DNA-PAINT probes may yield better images than standard probes in MINFLUX.

Post attachment
Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

Fluorophore flickering can create up to a tenfold increase in localization precision. We showed this can be mitigated by repeated scanning of a fluorophore.

Post attachmentPost attachment
Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

Starting off simple: misalignment of the phase pattern that generates a donut degrades performance, while misalignment of the pinhole has little effect.

Post attachmentPost attachment
Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

MINFLUX works well in theory, but system and sample imperfections can lead to suboptimal resolution. To address this, @JonasRies and I developed SimuFLUX: biorxiv.org/content/10.1101/…, a realistic simulator to optimize MINFLUX experiments before samples go on the microscope.

Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

Fluorophore flickering can create up to a tenfold increase in localization precision. We showed this can be mitigated by repeated scanning of a fluorophore.

Post attachmentPost attachment
Zach Marin avatar
Zach Marin@zacsimile·9 Apr 2025

Starting off simple: misalignment of the phase pattern that generates a donut degrades performance, while misalignment of the pinhole has little effect.

Post attachmentPost attachment
Reto Fiolka avatar
⟳ Zach Marin retweeted
Reto Fiolka@RetoPaul·2 Apr 2025

Our manuscript on accelerating OPM / SCAPE acquisition rate via aliasing is online now: opg.optica.org/boe/fulltext.… Our work addresses the issue that an OPM PSF is tilted, and as such it leaves a lot of "empty space" in Fourier space (FFT of experimental data on the right)... 1/n

Post attachmentPost attachment
石黑大雄 avatar
⟳ Zach Marin retweeted
石黑大雄@ppa007hot·28 Mar 2025

I am happy to share our 5-year's work of manual mapping of all the native proteins in the golgin family, providing the highest resolution atlas of the Golgi apparatus to date! Yes, we can distinguish proteins at different cisternae.

0:12
Abhishek Sau avatar
⟳ Zach Marin retweeted
Abhishek Sau@DrAbhishekSau1·19 Mar 2025

I’m excited to share our latest research from @musser_lab , published today in @Nature. Using two-color 3D MINFLUX, we directly visualized nuclear import and export events through transport-active nuclear pores. nature.com/articles/s41586-0…

Yiming Li avatar
⟳ Zach Marin retweeted
Yiming Li@YimingLi_SZ·21 Feb 2025

Excited to share LUNAR (Localization Using Neural-physics Adaptive Reconstruction) —a self-supervised, blind localization method that combines neural networks & physics to estimate aberrations from overlapped blinking molecules for large DOF imaging. Great work @judefffff !

0:27

Neural-physics adaptive reconstruction reveals 3D subcellular nanostructures over a large depth of field biorxiv.org/content/10.1101/… #biorxiv_biophys