Hi,
I am running analysis on RNA Seq data using DESeq2, I have two "treatments" for each sample each with two levels
status: sham or infected
temperature: high or low.
I have run the LRT to look at whether I am make sure I have not made any assumptions based on experimental design on whether genes are significantly dif expressed. To do this I have used the LRT
dstemp_lrt<-DESeq(dstemp, test="LRT", reduced= ~status)
res_LRT<-results(dstemp_lrt)
to give:
log2 fold change (MLE): temperature L vs H
LRT p-value: '~ status + temperature' vs '~ status'
Am I correct in thinking that this gives the Pvalues of the genes significantly differentially expressed as a result of whether it was sham or infected in comparison to those genes sig. differentially expressed when temperature is also factored in (the interaction)?
Therefore I should also run with "reduced=~temperature" to test the same but with just temperature and then compare the two lists?
Best wishes,
Rebekah
Ah cheers! So this can be used to identify if the DEG I have found when looking at the four individual experimental groups (infected L, infected H, control L, control H) can be attributed to the difference in temperature or difference in status?
Basically, yes. You should see which have the differences.