Trying to use calibrate3DsplinePSF but beads are not being identified

[oleland-br]

I'm trying to measure the PSF of beads in 3D using a cMOS camera. I saw people had issues with that before, but I've managed to manually adjust the default camera settings in the camera parameters because I keep getting a "could not get core metadata" error. I'll attach my metadata to ensure I'm setting the parameters correctly. Original Metadata - 1um-beads_488_003 - Denoised.txt

My problem lies in the fact that the software identifies random points instead of my fluorescent beads. I've adjusted all the settings in the GUI several times but find that the program never identifies the fluorescent beads. Perhaps this an issue in my camera settings, but I'm unsure where I'm going wrong.

Let me know if there's anything else I can provide to help troubleshoot.

[Rieslab]

Hello, My suspicion is that the threshold is too low, then random points are selected as beads and when they are too dense they are rejected. Please try increasing the ‘Relative Cutoff’ value. You can also increase the Filter size (it seems you have 20 nm pixel size, this is very small, we usually have 100 - 150 nm). So try filter size around 2-5 and a cutoff of 10 or even higher. For such small pixel sizes you would also have to increase the ROI size and the minimum distance. Alternatively, you could consider 4x binning during acquisition to get closer to the recommenced 100 - 150 nm pixel size. Good luck, Jonas

On 26. Jul 2023, at 22:57, oleland-br @.***> wrote:

I'm trying to measure the PSF of beads in 3D using a cMOS camera. I saw people had issues with that before, but I've managed to manually adjust the default camera settings in the camera parameters because I keep getting a "could not get core metadata" error. I'll attach my metadata to ensure I'm setting the parameters correctly. Original Metadata - 1um-beads_488_003 - Denoised.txt https://github.com/jries/SMAP/files/12177110/Original.Metadata.-.1um-beads_488_003.-.Denoised.txt https://user-images.githubusercontent.com/136110894/256362056-0a535067-f48a-4eb4-82d9-3fb0a525c098.png https://user-images.githubusercontent.com/136110894/256362112-da5e2f45-ee36-4a73-b3ea-3776ac9b32aa.png My problem lies in the fact that the software identifies random points instead of my fluorescent beads. I've adjusted all the settings in the GUI several times but find that the program never identifies the fluorescent beads. Perhaps this an issue in my camera settings, but I'm unsure where I'm going wrong.

https://user-images.githubusercontent.com/136110894/256361299-2a9fbe0e-5d5c-48ee-b173-e96a752560af.png https://user-images.githubusercontent.com/136110894/256361431-a57d562e-f69d-4ef8-9a5a-6d7a5ac892f9.png Let me know if there's anything else I can provide to help troubleshoot.

— Reply to this email directly, view it on GitHub https://github.com/jries/SMAP/issues/16, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZ52BSK5UN2DDCTD6OKBX3XSGACRANCNFSM6AAAAAA2ZEMJJQ. You are receiving this because you are subscribed to this thread.

---

Dr. Jonas Ries

• group leader -

Cell Biology and Biophysics Unit European Molecular Biology Laboratory Meyerhofstr. 1 69117 Heidelberg, Germany

Phone: +49 6221 3878199 Fax: +49 6221 387 8242 Mobile: +49 151 51040705 Email: @.*** http://www.embl.de/research/units/cbb/ries/index.html

[oleland-br]

I adjusted the parameters you suggested with a general setup like this:

The best identification I could get is illustrated below: I got something similar to this at a variety of settings, including filter: 5; cutoff: 10; min dist: 10; ROI: 15 filter: 5; cutoff: 10; min dist: 7; ROI: 15 filter: 5; cutoff: 10; min dist: 10; ROI: 15 filter: 5; cutoff: 12; min dist: 12; ROI: 15 filter: 2; cutoff: 12; min dist: 10; ROI: 15

Curiously, I got the error "Could not find and segment any bead. ROI size too large? " for the following settings: filter: 5; cutoff: 12; min dist: 12; ROI: 16 filter: 5; cutoff: 12; min dist: 12; ROI: 14 filter: 4; cutoff: 10; min dist: 10; ROI: 10 (amongst others, but really it seems like filter=5 was an important value to keep fixed)

For every bead calibration run, I got the same messages in the command window:

could not get core metadata roi could not be set SMAP image loader used

The first is unsurprising to me, but I fixed the default camera values to be identical to my camera settings. I'm not sure what the second one means.

To note, the voxel size is 0.0232x0.0232x0.1um and images are acquired with an sCMOS camera. The setup for the bead image acquisition is identical to our experimental setup for microbe imaging so we don't want to alter the PSF with binning. Do you have any other suggestions on how to fix the bead segmentation issue?

[Rieslab]

It seems that now hardly any beads were recognised (they lack the magenta marks). Try lowering the cutoff again. If you like, you can also share the bead tiff files with me, then I can give it a try and tell you working parameters. Ignore the messages on the loader and metadata. The 3D calibration does not use any of those (units are in pixels, the PSF is normalised). Best, Jonas

On 27. Jul 2023, at 22:58, oleland-br @.***> wrote:

I adjusted the parameters you suggested with a general setup like this: https://user-images.githubusercontent.com/136110894/256661202-eba4da77-bcdb-4e62-8ec5-5c3614d80c7c.png The best identification I could get is illustrated below: https://user-images.githubusercontent.com/136110894/256661393-532d48c0-ba5d-453d-a9fc-ed089f64f4ce.png I got something similar to this at a variety of settings, including filter: 5; cutoff: 10; min dist: 10; ROI: 15 filter: 5; cutoff: 10; min dist: 7; ROI: 15 filter: 5; cutoff: 10; min dist: 10; ROI: 15 filter: 5; cutoff: 12; min dist: 12; ROI: 15 filter: 2; cutoff: 12; min dist: 10; ROI: 15

Curiously, I got the error "Could not find and segment any bead. ROI size too large? " for the following settings: filter: 5; cutoff: 12; min dist: 12; ROI: 16 filter: 5; cutoff: 12; min dist: 12; ROI: 14 filter: 4; cutoff: 10; min dist: 10; ROI: 10 (amongst others, but really it seems like filter=5 was an important value to keep fixed)

For every bead calibration run, I got the same messages in the command window:

could not get core metadata roi could not be set SMAP image loader used

The first is unsurprising to me, but I fixed the default camera values to be identical to my camera settings. I'm not sure what the second one means.

To note, the voxel size is 0.0232x0.0232x0.1um and images are acquired with an sCMOS camera. The setup for the bead image acquisition is identical to our experimental setup for microbe imaging so we don't want to alter the PSF with binning. Do you have any other suggestions on how to fix the bead segmentation issue?

— Reply to this email directly, view it on GitHub https://github.com/jries/SMAP/issues/16#issuecomment-1654570417, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZ52BV4MZTN7WML6INRN4DXSLI6BANCNFSM6AAAAAA2ZEMJJQ. You are receiving this because you commented.

---

Dr. Jonas Ries

• group leader -

Cell Biology and Biophysics Unit European Molecular Biology Laboratory Meyerhofstr. 1 69117 Heidelberg, Germany

Phone: +49 6221 3878199 Fax: +49 6221 387 8242 Mobile: +49 151 51040705 Email: @.*** http://www.embl.de/research/units/cbb/ries/index.html

[oleland-br]

I emailed over one of the tif files with the metadata. Thanks!