Hi !
I have a RNAseq samples with this design:
sample genotype traitement
Col_1 wt normal
Col_2 wt normal
Col_3. wt normal
ColHS1 wt heatshock
ColHS2, wt heatshock
ColHS3 wt heatshock
hon4_1 hon4 normal
hon4_2 hon4 normal
hon4_3. hon4 normal
HS1 hon4 heatshock
HS2. hon4 heatshock
HS3. hon4 heatshock
I'm only interest to the genes differently expressed between my two genotypes, so I make this design :
design= ~ genotype + traitement
and then I retrieve my genes of interest in this way:
results(dds,contrast = c("genotype","hon4","wt"))
I the protocol good ?
I have 600 DEG with this method, are all the genes specific to the mutation ?
I mean, there are no genes in this list that could be related to the interaction with the treatment ?
Thank's by advance for you're help, I read help with ?results but I have a little trouble with the experimental designs
I had the interaction term on the design formula and so I have multiple set of genes:
DEG for the effect of treatment in my WT DEG for the effect of treatment in my mutant DEG for diff between WT and mutant (no treatment) DEG for diff between WT and mutant (treatment) and finally: DEG for the different response to treatment in genotypes (interaction)
I do not see how really how to approach this notion of interaction, I have around 500 genes deregulated for the interaction term, it would mean that the mutation has a greater impact in case of stress ? (because if I had 0 term, that would have meant that the mutation is independent of the treatment, If I understood correctly)
Thank you for you're time, I'm bioinformatic student and I really need to improve my statistics skills aha