Any reference/paper that might have used such an approach?
In most cell line experiments, it is widespread to perform RNA-seq for many cell-lines before or after a treatment (of a drug). Then, typically, differential expression analysis is conducted.
Now if I have a pre-treatment expression of a new cell-line, how to determine what will be its likely (the drug) post-treatment expression? Assume drug efficacy and response are the same across all cell lines.
1 answer
You would have to build some form of a predictive / regression model using reference datasets consisting of pre- and post-treatment cell-lines, and then apply this model to the new cell-line in order to infer response (within a certain window of statistical error). How you define 'response' is your role, or that of the clinicians / biologists working with you.
Otherwise, if you know your drug and its target(s) quite well, then you may simply have to examine the expression of key target genes in your sample. For example, Trastuzumab (Herceptin®) targets ERBB2 / HER2; so, the cell-line would have to express this in order to respond to the drug. By simply examining normalised counts or Z-transformed expression data, one could infer heightened expression of a particular gene or genes.
Kevin
Log in to answer this question.