I have some queries regarding differential expression using deseq. I have time-point data with 3 biological replicates and control with biological replicates.
I have generated the counts using htseq-count. I want to be able to compare my time points with control. My data is like following
sampleName fileName condition Control1.count Control1.count Control Control2.count Control2.count Control Control3.count Control3.count Control 3hr_rep1.count 3hr_rep1.count 3 hr 3hr_rep2.count 3hr_rep2.count 3 hr 3hr_rep3.count 3hr_rep3.count 3 hr 6hr_rep1.count 6hr_rep1.count 6 hr 6hr_rep2.count 6hr_rep2.count 6 hr 6hr_rep3.count 6hr_rep3.count 6 hr 12hr_rep1.count 12hr_rep1.count 12 hr 12hr_rep2.count 12hr_rep2.count 12 hr 12hr_rep3.count 12hr_rep3.count 12 hr samples <- read.table(file.path(directory, "control_sample"), header = TRUE) dds <- DESeqDataSetFromHTSeqCount(sampleTable = samples, directory= directory, design= ~condition) dds$condition<- relevel(dds$condition, "Control") dds <- DESeq(dds) res <- results(dds)
I have a few questions.
- How to take care of biological replicates to identify differentially expressed genes between conditions.
- How can I design dds so as to compare all my time points with respect to control. -Also, is it possible to a comparison between all sample at once.
I want to be able to produce a heatmap with each column representing data from three biological replicates if that makes sense. I am presently able to plot normalised read counts where each replicates is represented as distinct sample and not considering the replicates aspect.
I want help in designing the dds. Any suggestion is appreciated.
Many thanks in advance.