I'm performing statistical analyses on gene expression data. My samples have two different conditions. The conditions are about the genotype and health of the plants. I have 4 different situations: Genotype1-Healthy, Genotype2-Healthy, Genotype1-infected, Genotype2-infected.
I would like to evaluate the statistical differences of the expression values of all combinations of the groups. They will be six melds total.
I've already done the two-way anova:
Myanova<-aov(Value~Condition1*Condition2,data=GENE).
This is for example the output for the values of an analysed gene.
Df Sum Sq Mean Sq F value Pr(>F)
Condition1 1 0.76131 0.76131 5.7649 0.035169 *
Condition2 1 2.53138 2.53138 19.1684 0.001102 **
Condition1:Condition2 1 0.01125 0.01125 0.0852 0.775811
Residuals 11 1.45267 0.13206
I would like to continue with the post-hoc bonferroni test.
I searched on the net but I couldn't find the right function that gives me the values for the six different combinations.
I tried the parwisettest but that's not what I'm looking for:
pairwise.t.test(Gene$Value,Gene$Condition2, p.adj="bonferroni")
pairwise.t.test(Gene$Value,Gene$Condition1, p.adj="bonferroni")
Could you tell me how to easily perform the dunn-bonferroni post-hoc test for the multiple comparison of my samples?
