Hello around,
I got a question on genome-wide gene expression microarrays (e.g. Agilent): When filtering the data on flags there are two options: using detected or not detected flags, or both (assuming compromised flags are of too low quality to be taken). If planning to have reliable, robust, reproducible data to carry on for further experiments (e.g. qPCR, NGS etc.), or for publishing results, are “not detected” signals too bad?