Hi,
I have some RNAseq data derived from monkeys that have been treated with either drug A or drug B and sampled in time series.
In total there are 35 samples:
7 monkeys (3x group A, 4x group B), sampled 5 times each on 5 different days.
The challenge I am facing is that it appears my samples are clustering by individual (monkey ID, see the PCA below). I felt that the way to get around this is to treat each individual as a batch.
Day 0 represents the baseline measurement, before treatment. I've included a PCA on the data below, you'll notice that the baseline measurements don't cluster together at all, which is expected for human and monkey data - the intrinsic differences between individuals is often much greater than the treatment effect.
It is for this reason that I'd like to compare each point in the time series to each individual's baseline. So, I would like to get DE results comparing DX/D0, per animal if it is possible. Or if not possible, I would like to be able to model DX/D0 for each timepoint and treatment, however by taking into account the apparent batch in my data (ex: only comparing D1/D0 within each animal, and then getting a gene list that represents the DE for all the animals in that treatment group). I honestly have no idea if this is even possible, so if anyone has suggestions on how to proceed with this type of data I'd love to hear!! :)
....thanks for the quick reply. I'll look closer at the documentation next time.