search

hci-unihd / antibodies-analysis-issues

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

QC: Minimal number of control cells #81

Open tischi opened 4 years ago

tischi commented 4 years ago

Lets plot histograms of control cell numbers per well and see what is the lowest number we get and than check those wells and the values we get in them to see if something is weird... that should give us a clue of the good value there

tischi commented 4 years ago

Number of control cells per well

image

tischi commented 4 years ago

IgG robust z-score vs number of cells per well

image

tischi commented 4 years ago

@metavibor @constantinpape See above plots for an overview about the number of control cells per well. I made the lower plot to see whether a too low number of control cells would lead our scores to become unstable. Maybe for plate 413 this could be the case... Any ideas how to determine a threshold from this? Currently not obvious to me.

tischi commented 4 years ago

By mutual agreement we defined 100 as the threshold.