search

hatzakislab / DeepFRET-GUI

GNU General Public License v3.0
8 stars 5 forks source link

Calibrate Alignment of Green / Red Channel #26

Open BradyAJohnston opened 3 years ago

BradyAJohnston commented 3 years ago

I'm wondering if there is a way to calibrate the left / right channels - given that during the experiment they were not perfectly symmetrical. In other software that we were previously using for this, we could calibrate with a standard - but it seems that deepFRET just divides the sensor in half and overlays them without any possibility for adjustments?

Example image below, with highlighted spots that are the more obvious examples of spots that should align circled in pink. image

Because of this off-alignment, the spots that do 'co-localise' are just jibberish as the signals belong to different molecules.

Any chance that this is possible with deepFRET? The raw data is just .tif movies that look like below.

5min_7_MMStack_Pos0ome_ave

graciesiffer commented 1 year ago

We are having the exact same problem, did you ever find a solution?

BradyAJohnston commented 1 year ago

@graciesiffer sadly not! I ended up abandoning this approach as it was unusable for this type of data in the current state.