search

hci-unihd / antibodies-analysis-issues

Issue tracker for problems in the antibodies analysis workflow.
0 stars 0 forks source link

Plate titration_plate_20200403_154849 Site G07-0004 #19

Open tischi opened 4 years ago

tischi commented 4 years ago

{"plateName":"titration_plate_20200403_154849","siteName":"G07-0004","pixelLocation":[19775.419038860786,17835.182072682368,0.0]}

The marker channel is partially screwed up in this well. Maybe something with the microscope? We should ask Vibor.

tischi commented 4 years ago

image

constantinpape commented 4 years ago

FYI, @wolny has set up an outlier detection pipeline and a student of Vibor has used it to mark outliers for several plates already. It's not used in the analysis yet, pending https://github.com/hci-unihd/batchlib/pull/34. I will have a look at the PR now and try to integrate it. In the longer term we can think about:

(we can make issues in batchlib to discuss this more generally)

tischi commented 4 years ago

But we have to be careful not to exclude interesting biology. Maybe in those wells there was really less virus...

tischi commented 4 years ago

@metavibor If you look at the screenshot above, could you comment on

  1. why you would think some of the images in this well have a lower virus signal than the other images
  2. in general, why is the signal to background ratio in the virus channel so low? Is there fluorescence in the medium?