Hello everyone,
There is a matter that I need your help and knowledge. Currently, I am analyzing some interesting microarray and RNA-seq datasets for my PhD project. There is a dataset which has 4 cell lines, and in each cell line there are 4 conditions as, Control, TNFa, IFNa, and IFNg treated. For each condition there is one biological replicate. What I would like to do is perform DEGs analysis between Cell Line X conditions CTRL vs TNFa. What sort of approach would you recommend me? Since there is no at least 2 biological replicates, I cannot apply algorithms like limma.
As a solution first I thought that maybe merging the same conditions across multiple cell lines would be an alternative. However, in the CTRL groups of all cell lines, the baseline expression of MITF gene is not similar which is important for us since we determine the differentiation state of the cell line by looking at the expression of MITF gene.
I am looking forward to receive your comments and learn from you! Thank you.
You can do
~condition + cell_lineto eliminate the massively confounding effect of different cell lines and with this test the condition effect. That of course assumes that the condition effect is somewhat similar.Thank you very much for your valuable reply! I got very nice DEGs for TNFa and IFNg groups which totally makes sense. However, for IFNa group I did not any DE gene which has p.adj.val < 0.05. I believe that this might arise from the lack of biological replicates, and potentially IFNa might have a weaker or different effect compared to TNFa and IFNg.
Thanks again!