I have time series data from two gene background:

    1)  if I want to compared 0 hour first in each background and then see how is the difference between two gene background at all other time point(1h and 4h which already removed the effect of 0h), in other word, I do not want to know the different at 0h now, just want to see the fold change compared with 0h difference between the two background at all time point( 1h and 4h).  I design like below; I am not sure I am doing right or not.

dds <- DESeqDataSetFromMatrix(countData = countData, colData = colData, design = ~Time+Treat+Time:Treat)

dds<- DESeq(dds, test="LRT", reduced = ~ Time+Treat)

2) And if I want to see all time point different including 0 h (means no normalization 0h first) between two background, I design like this:

dataset<- DESeqDataSetFromMatrix(countData = countData, colData = colData, design = ~Time+Treat+Time:Treat)

dds <- DESeq(dataset)

About the two conditions above, I am doing right thing or not? I am very confusing about how to use the reduced option. Thank you for helping.

On the support site, I provide guidance on particular software related questions, but unfortunately I don’t have time to provide general statistical consulting. It’s just not feasible for me to do this. Beyond the example and documentation, I’d again recommend you work with a statistical or bioinformatics group on designing your analysis.