I have been reading about drug-gene interactions and pharmacogenetic tests. I haven't been able to find many resources that define the process of testing for possible drug-gene interactions without performing trial and error method. For example, drug-drug interactions can be predicted by using several parameters as noted here. Are there any parameters similar to the ones described in given paper for predicting drug-gene interactions if drug characteristics and genetic data is available? Please describe process or provide source.
One can list candidate genes based on their functions.
If a drug acts by inhibiting the expression of gene
G in order to reduce process
P, then, gene
G, all its regulatory elements and all sequences involve in process
P are good candidates for such interaction and should be investigated. If there is known genetic variation in the population at these loci and if we know this genetic variation affect phenotypic variance, then definitely, those are good candidates.
However, to my knowledge, this is the extend at which we are able to go. We can list candidates but it does not mean in the absolute that a candidate must present an interaction with the drug of interest and it does not mean that a non candidate gene would necessarily not have an interaction. Only a statistical testing for such interaction would allow one to solve the mystery.