We have an RNA-seq time-series over three time points, measuring the response of a cell line to three treatments (A1, A2 and B). We also have the control measurements, over the same time points plus the time 0h (treatments were applied after 0h). We have three biological replicates for each time point and treatment.
Treat Time 1 Ctrl 0h 2 Ctrl 0h 3 Ctrl 0h 4 Ctrl 1h 5 Ctrl 1h 6 Ctrl 1h 7 Ctrl 2h 8 Ctrl 2h 9 Ctrl 2h 10 Ctrl 3h 11 Ctrl 3h 12 Ctrl 3h 13 B 1h 14 B 1h 15 B 1h 16 B 2h 17 B 2h 18 B 2h 19 B 3h 20 B 3h 21 B 3h 22 A2 1h 23 A2 1h 24 A2 1h 25 A2 2h 26 A2 2h 27 A2 2h 28 A2 3h 29 A2 3h 30 A2 3h 31 A1 1h 32 A1 1h 33 A1 1h 34 A1 2h 35 A1 2h 36 A1 2h 37 A1 3h 38 A1 3h 39 A1 3h
We would like to identify the genes that respond to treatments A1 and A2, but not to B. Treatments A1 and A2 might have different dynamics, so we cannot merge their labels in a single factor. What would be the best way respond to this question with edgeR? For now, we are not interested in highlighting different response dynamics.
Thank you in advance,