Identification of cancer-associated promoters

[alejandrogzi]

Hi @jleechung,

I am trying to reproduce part of the proActiv paper by combining different cancer/normal RNA-seq datasets. To identify cancer or pan-cancer associated promoters a batch effect removal is necessary; however, I am not quite sure how/when to apply limma using proActiv. Does integrateProactiv() function renormalization solve this?

Second, the linear models used in the paper to identify these promoters are used during proActiv default calculations or need to be specified somewhere?

Thanks in advanced, Alejandro