Hello all, I have a time course and 2 groups (Control and Exercise), however, since my control is only one group I think I have a nested relation between groups and time and then I can't perform the traditional design as proposed in the vignette (full: Groups+Time+Groups:Time vs Reduced: Groups + Time). I'm wondering how could I best design the matrix to compare the control with the exercise groups.

Thanks in advance

Samples Groups  Time
90001017003 Control 0h
90020017003 Control 0h
90037017003 Control 0h
90038017003 Control 0h
90109017003 Control 0h
90110017003 Control 0h
90127017003 Control 0h
90128017003 Control 0h
90145017003 Control 0h
90146017003 Control 0h
90025017003 Exercise    0.5h
90026017003 Exercise    0.5h
90043017003 Exercise    0.5h
90044017003 Exercise    0.5h
90115017003 Exercise    0.5h
90133017003 Exercise    0.5h
90134017003 Exercise    0.5h
90152017003 Exercise    0.5h
90007017003 Exercise    0.5h
90008017003 Exercise    0.5h
90027017003 Exercise    0h
90028017003 Exercise    0h
90045017003 Exercise    0h
90046017003 Exercise    0h
90117017003 Exercise    0h
90118017003 Exercise    0h
90135017003 Exercise    0h
90136017003 Exercise    0h
90009017003 Exercise    0h
90010017003 Exercise    0h
90029017003 Exercise    1h
90047017003 Exercise    1h
90048017003 Exercise    1h
90119017003 Exercise    1h
90120017003 Exercise    1h
90138017003 Exercise    1h
90155017003 Exercise    1h
90156017003 Exercise    1h
90011017003 Exercise    1h
90012017003 Exercise    1h
90039017003 Exercise    24h
90040017003 Exercise    24h
90112017003 Exercise    24h
90129017003 Exercise    24h
90147017003 Exercise    24h
90148017003 Exercise    24h
90023017003 Exercise    48h
90041017003 Exercise    48h
90042017003 Exercise    48h
90114017003 Exercise    48h
90150017003 Exercise    48h
90005017003 Exercise    48h
90013017003 Exercise    4h
90014017003 Exercise    4h
90031017003 Exercise    4h
90032017003 Exercise    4h
90050017003 Exercise    4h
90121017003 Exercise    4h
90122017003 Exercise    4h
90139017003 Exercise    4h
90140017003 Exercise    4h
90157017003 Exercise    4h
90017017003 Exercise    7h
90018017003 Exercise    7h
90053017003 Exercise    7h
90126017003 Exercise    7h
90143017003 Exercise    7h
90144017003 Exercise    7h

This isn't a Bioconductor question per se, but instead a basic stats or experimental design question. So biostars might be a better venue to ask questions.

That said, what's the difference between the controls and the exercisers at time zero? Are the exercisers at different times the same subjects? I see this type of experiment from time to time and it always seems to me to be a waste of money (running the controls, that is). Others may have differing opinions, but I personally don't see a real use for the controls, unless they are supposed to represent those who have never exercised and the exercisers at time zero are like super buffed and are just about to get their exercise on.

And even then, what's the point? It seems like you have a time course following people up to 7 hours after exercising and might want to know what genes are changing over that span of time. But if that's true you need to block on subject, and the controls are now just sitting there looking useless, because they are completely different people.

Anyway, like I said, this isn't really a Bioconductor question, and experimental design and analysis questions are better asked of someone local who knows what's up statistically rather than some randos on the interwebs. But if you like the rando approach, maybe try biostars.