we found that a threshold of FC > 4 was necessary to call a gene reliably upregulated (type S error <5%;Figure 1B), in line with MAQC/SEQC conclusions. In contrast, estimating FC from a higher number of datasets dramatically improved replicability: for 2 datasets the 5% error threshold i...
While there are no rules you have to follow, there are a few things to keep in mind: ◆【5】 If you aim too high at the beginning, you may never get the site off the ground. You can always add something new to your site.