A reviewer asked us to perform some form of quality control to filter profiles prior to analysis.
This is not a typical step in the image-based profiling pipeline, but it is important to compare metrics if we perform some QC. It does not make sense to filter profiles based on cell count (see #63), but it does make sense to filter profiles in
edge wells given our well position effect observations.
In this PR, I calculate percent strong after filtering edge wells in both assays. I add two rows and two bars to the following supplementary figure:
A reviewer asked us to perform some form of quality control to filter profiles prior to analysis.
This is not a typical step in the image-based profiling pipeline, but it is important to compare metrics if we perform some QC. It does not make sense to filter profiles based on cell count (see #63), but it does make sense to filter profiles in edge wells given our well position effect observations.
In this PR, I calculate percent strong after filtering edge wells in both assays. I add two rows and two bars to the following supplementary figure: