Hi,

I would like to know if there is any Bioconductor R package that can deal with RNA-seq time-series allowing for time offset between individuals. That is, I have 10 individuals, each with variable sampling (variable in both sampling time and number of samplings) throughout their pregnancy and I would like to have those genes that show a significant difference in expression throughout pregnancy. A mock design example is below:

INDIVIDUAL Sampling_day

Individual1 41

Individual1 80

Individual1 120

Individual1 150

Individual1 200

Individual2 30

Individual2 70

Individual2 180

Thanks,

João

This seems fairly standard. Just use a design with individual and time effects; the usual suspects (edgeR, DEseq2, limma) are all able to deal with this. For example, in edgeR, one might define the design matrix as:

design <- model.matrix(~individual + time)

... and then test the effect of the final term. If you want to account for non-linear trends, you can use a spline basis matrix. There should be advice in the relevant sections of the user guides on how to do this.

The fact that there is a offset in time between individuals is largely irrelevant, as the time factor is still mostly orthogonal to the individual factor (at least for the first two individuals you've shown above). You would have more problems if all early time points belonged to one individual and all late time points belonged to another individual.