Dynamic/timeline RNAseq experiment design, reps vs time-points?

[MikhailFokin]

Hi all, I am planning the time-course RNAseq study. Obviously, I am interested in max number of time-points, but I am limited by number of samples I can afford

So, options are
A. 3 biol reps X 2 time-points
B. 2 biol reps X 3 time-points
C. 1 biol reps X 6 time-points

Some extra info:
- I am not planning "Nature" article... thx for applause it is a kind of in-house experiment
- I am interested, first of all, in a few dozens of genes, so I can test some of them by qPCR to be sure that my biol reps are identical (for each time-point) before NGS. I've studied the same kind of model before and biol reps were 99.999% identical in terms of expression

Option A is traditional, but I think (and I am biased towards) that with option C I will be able to use something like genome-scale correlation analysis to group genes into co-expressed cohorts much more effectively than in var.A

Any strong cons against option C?

PS there were couple of similar questions before... with no answers

[luc]

What organism? How closely related are the individuals?

[MikhailFokin]

thx luc!
The organism - diploid mold, ~2x10k genes of which I am interested in ~100 highly expressed specific genes.

Biol reps - are from exactly the same strain, the same inoculum. So they are identical (except harvesting time i.e. growth phase).
The main goal is to investigate dynamics and correlations between different genes in time (in case of any concerns I can check more reps by qPCR).

[MikhailFokin]

technically with option C (1 rep x 6 times), I will have multiple (by number of genes) vectors of length 6. And will be able to group them into clusters by absolute or relative expression

[luc]

Hi Mikhail,
I would suggest to have your cake and eat it too (three replicates and six time points) by using a cheaper RNA-seq protocol. Assuming you have a reference genome and gene annotations, 3'-Tag-Seq should give you all the data needed. You could use for example kits from Amaryllis Nucleics or Lexogen.

[MikhailFokin]

Thank you! Interesting options... Since there is clearly the gap between qPCR and RNAseq.