Does anyone know if the q- and p-values output by CuffDiff are reliable when more than two treatments are compared in the the analysis.
Generating q-values for each gene in a two-treatment study is fairly strait forward. However, in a five-treatment study where each pair of treatments is compared, there are 15 pairwise comparisons for each gene. Can these comparisons be lumped together to generate the q-values, or should each pair of treatments be analyzed independently?
Generating q-values for each gene in a two-treatment study is fairly strait forward. However, in a five-treatment study where each pair of treatments is compared, there are 15 pairwise comparisons for each gene. Can these comparisons be lumped together to generate the q-values, or should each pair of treatments be analyzed independently?
Comment