In my opinion, you may be ok... IF your control is identical for all conditions; in your case, I assume (but I could be wrong), your control is a sample at time-zero, and then you collected from 1hr, then again at 6hr. You would repeat the same 0hr sample on both plates. If there is inter-plate variability, that should be reflected in the different Cts of the control, which would calibrate the two later time points. This seems reasonable.

If, however, you did the experiment on two separate occasions, collecting 0hr and 1hr, and then on some other day a different 0hr and 6hr, you would need the intercalibrator to help standardize the data. You may also find this VanDeSompele paper helpful: doi:10.1016/j.bdq.2015.07.001

I had a similar issue, and my solution was to test all genes and each condition (and their controls) on a single plate for one biological replicate, and test different bio-reps on different plates - so the layout of each plate was the same (example: in your case, a single plate would hold 0hr, 1hr, 6hr vs geneX, geneY, geneZ, reference geneA, reference geneB, reference geneC; plate 1 = bio-rep1, plate 2 = bio-rep2, plate 3 = bio-rep3). When I calculate relative expression, I use a modified 2^ddCt method (pfaffl) that can incorporate different efficiencies for each amplicon, but it's essentially the same equation. I calculate RE of each bio-rep individually, and then find their average (and standard error). Some people may take issue with this, but I believe my calculation resolves technical variability because the proper controls are matched within the plate. 

You should also consider which is more important to you: the difference in expression between multiple genes (geneX is up, geneY is down) for any condition, or the difference in expression between multiple conditions (1hr is up, 6hr is more up) when designing your plate layout. If the first case is more important, you should run as many genes as possible for one sample and its control partner in a single plate. If the second case is more important, you should run as many conditions and controls in a single plate, so their relationship is clear, but test different genes on separate plates. And, of course, you have to maximize all of this so that you are being efficient with sample and reagent. It's a difficult balance. Best of luck!

Hi, I always wondered the same thing. I don't think we can afford and interplate calibrator. Is it reasonable to use a sample mix in all runs and calibrate it that way?

In my opinion, you may be ok... IF your control is identical for all conditions; in your case, I assume (but I could be wrong), your control is a sample at time-zero, and then you collected from 1hr, then again at 6hr. You would repeat the same 0hr sample on both plates. If there is inter-plate variability, that should be reflected in the different Cts of the control, which would calibrate the two later time points. This seems reasonable.

If, however, you did the experiment on two separate occasions, collecting 0hr and 1hr, and then on some other day a different 0hr and 6hr, you would need the intercalibrator to help standardize the data. You may also find this VanDeSompele paper helpful: doi:10.1016/j.bdq.2015.07.001

I had a similar issue, and my solution was to test all genes and each condition (and their controls) on a single plate for one biological replicate, and test different bio-reps on different plates - so the layout of each plate was the same (example: in your case, a single plate would hold 0hr, 1hr, 6hr vs geneX, geneY, geneZ, reference geneA, reference geneB, reference geneC; plate 1 = bio-rep1, plate 2 = bio-rep2, plate 3 = bio-rep3). When I calculate relative expression, I use a modified 2^ddCt method (pfaffl) that can incorporate different efficiencies for each amplicon, but it's essentially the same equation. I calculate RE of each bio-rep individually, and then find their average (and standard error). Some people may take issue with this, but I believe my calculation resolves technical variability because the proper controls are matched within the plate. 

You should also consider which is more important to you: the difference in expression between multiple genes (geneX is up, geneY is down) for any condition, or the difference in expression between multiple conditions (1hr is up, 6hr is more up) when designing your plate layout. If the first case is more important, you should run as many genes as possible for one sample and its control partner in a single plate. If the second case is more important, you should run as many conditions and controls in a single plate, so their relationship is clear, but test different genes on separate plates. And, of course, you have to maximize all of this so that you are being efficient with sample and reagent. It's a difficult balance. Best of luck!

Hi i am really grasping on straw horse here but i have been asking this below question and researching online but to no avail. can anybody answer this below?

Question: Time course gene expression analyses: RNA seq over mass spec?

why when wanting to look at genes that respond at say 1hr (early response), 6hr(intermediate response), 16hr (late response) it is better to look at RNA expression rather than protein?

has it got something to do with RNA instability?

Can someone provide me with the formula to calculate interplate calibration, the way Genex does it (two references samples that are on plate 1 and plate 2)?