Exp Design
DESeq2_1.22.2
I am in the processes of performing an RNA-Seq analysis on samples collected from a treatment conducted at day 21 and day 35. I have two groups, control and treated, and in those two groups, they are split into day 21 and day 35. I have 3x replicates for both.
Sample | Treatments | day |
---|---|---|
G1 | control | 21 |
G1 | control | 21 |
G1 | control | 21 |
G2 | drug | 21 |
G2 | drug | 21 |
G2 | drug | 21 |
G1 | control | 35 |
G1 | control | 35 |
G1 | control | 35 |
G2 | drug | 35 |
G2 | drug | 35 |
G2 | drug | 35 |
Goal
I would like to compare control vs. drug at time points 21 and 35 days. As well as drug at 21 vs drug at 35 days.
Code
dds <- DESeqDataSetFromMatrix(countData=counts(expSet), colData=pData(expSet), design= ~day + treatments)
dds <- DESeq(dds)
resultsNames(dds)
This returns
[1] "Intercept" "day_35_vs_21" "treatments_drug_vs_control"
Running results and lfcShrink
res <- results(dds, contrast=c('treatments', 'drug', 'control'), alpha=0.05, lfcThreshold=0.585)
lfcRes <- lfcShrink(dds, coef='treatments_drug_vs_control', res=res, type='apeglm')
Question
From my understanding, the design here is in layman's terms, taking into account days effect on treatment, what is the difference in treatment vs control. Is this correct?
Also, would the day_35_vs_21
, therefore, mean that taking into account treatments effect on the day what is the difference at day 21 vs day 35.
What would the design have to be if I wanted to see day 21 drug vs day 35 drug or control at day 35 vs drug at 35? Would I need to do something along the lines of day:treatment or would I need to do:
dds$group <- factor(paste0(dds$day, dds$treatments))
deisgn(dds) <- ~group
Thank you in advance for your help!