I have downloaded the 19Q4 PRISM primary screen collapsed logfold change file. I am noticing that for some cell lines majority of the drug values are NA, is it right to assume that NA values are because they have been filtered out for some reason? When focusing on just breast cancer cell lines, it seems like majority of the cell lines have NA values across most of the drugs. Is the correct interpretation that these breast cancer cell lines did not respond well or are not suitable for drug treatments?
Apologies for the late response. The NA’s are essentially assay failures and redacted across the compound plates so the failures are not really drug dependent. For a detailed description of the QC process please refer to the methods part of Discovering the anticancer potential of non-oncology drugs by systematic viability profiling | Nature Cancer . For the sake of interpretation, the mode of failure in the assay was most relevant to the growing those cell lines in the assigned pools in that particular batch, so there might be many reasons why they failed in that particular assay (might be some barcoding issue, some very fast cell line in the pool taking over the pool, etc.) and I would be hesitant to conclude those cell lines are not suitable for drug treatments.
Also, in the next iterations of the PRISM screen, we recovered most of the breast cell lines and therefore I would assume they should be fine for future screens.