Dear all I have resepctively performed linear regression analyses on two idependent datasets. One is conducted on log2 transformed CPM from RNA seq, while another is on log2 transformed scaling data from micro-array. I want to combine the effect of genes and generate meta pvalues using metafor package. The units of the independent variable were the same in two dataset. But I am concerned that the dependent variable (gene expression) were not at a comparable level, whcih will induce different effect, and make the meta effect and pvlue unreliable. What pre-processing I can do to deal with this.

Best regards.