Our lab is working on improving our RNAseq pipeline, specifically when it comes to filtering out low-quality genes. We are currently using a global RPM cutoff. All members of any 1 biological group within an experiment must meet that threshold or the gene is filtered out. We are worried that no matter what we do, this almost always removes smaller genes because they intrinsically have a lower number of reads. We'll continue to pro-con in our lab for the foreseeable future, but I want to know how other researchers are accommodating this conundrum of "What defines a useless gene in our analysis?"

I think I've conveyed everything to cover rule 2.

View Poll