I am looking at the effect of stress on a number of stress genes via a qRT PCR. I have put them into graphpad where my groups are +/- the chemical inducing stress and each row is one gene - all the genes are tested from the same sample. I am now carrying out a 2-way ANOVA test to see if there is any difference in the genes upon adding stress. As these are individual genes that do not effect each other and since I am only making a row-wise mulitple comparison should I use Fischer's LSD test? Each row gene has been analysed individually (as I use fold changes) and so there is interaction per column per say except for the fact that each column is from the same sample (I have remembered to match the column for the analysis). But should I use the Tukey's or the Ficher's LSD test on GraphPad?