Issue loading my UCLA Miniscope V4 recordings into MiniAn

[martinagkr]

Hi all,

I am quite new to using Miniscope processing pipelines and just working with Miniscope recordings in general. I have used MIN1PIPE a bit but found that for my data (some mice have far less expression than others and generally I am working with low SNR recordings) having the option to tune parameters via interactive visualisations might be a more ideal option.

Thus I am trying to get familiar with miniAn, I ran through the notebook pipeline with the demo data with no issues - so nothing seems to be wrong with my installation/dependencies. However, when I have tried to run the pipeline on my own data at the very start in the raw visualisation step my data does not look like it is accurately represented. I have ran the same data through MIN1PIPE and therefore have checked that it's not an imaging problem directly. I think there might be some setting that I would have to adjust like a thresholding issue but i'm coming up blank and have been stuck for some time. I have recorded at 10 FPS, LED:14-18 and the recordings i'm loading are uncompressed (compression: GREY in user config file).

As such, the videos are quite large 0.7 GB each (11 per session) and despite adjusting the memory limit the pipeline only seems to load 3 videos out of 11. The videos in the demo data file are about 10x smaller so this is one big discrepancy but unsure what else may be so different that the raw visualisation is virtually null (mean fluorescence plot is near 0)

I have attached a screenshot in case I am not explaining this well: