Not to mention that recent studies show that the 'normal' adjacent tissue of a tumor reflects more of a tumor environment, rather than normal tissue environment. Can i still use the adjacent tissue as my reference sample?

https://www.nature.com/articles/s41467-017-01027-z#Sec10

To your second question, that's something you will have to decide for yourself.

To your former question, comparing two sets of values is really all you ever do in qPCR, so I don't see why this should be problematic.

You have Cq values for miR-A in both tissues (high macrophage, low macrophage), you have Cq values for U6 in both tissues. Assuming U6 is an appropriate reference, then for each sample you just calculate the dCt (Cq miR-A - Cq U6).

dCt values represent your normalised expression level, and are entirely arbitrary in terms of absolute amounts (lots vs little) but are all you need to comparing relative amounts (more vs less).

And then you run stats between all the dCts for one tissue type vs the other tissue type. dCts are approximately normally distributed, so most statistical tests should be ok. For two groups, you could just use a T-test.

Just remember that this is qPCR, so low values represent higher expression: this applies both before and after normalisation. If you want to plot your data in a more intuitive fashion, you can just invert the sign (-dCt).

hi John Hildyard tq so much for your prompt answer. Since i am only comparing the amount, i don't really need a calibrator/ reference sample right? I could skip the ΔΔCt step and just stop at ΔCt. But to calculate the level of relative gene expression or fold change, shouldn't I still need ΔΔCt for the Livak formula? If i were to stop only at ΔCt, what shall I call this method since it's not actually the comparative Ct method? My concern is that I have not come across any papers that only compare relative amounts of 2 different groups. I am sorry to ask so many questions, John. I have gone through many papers and have asked around as well, but they all seem to have a control/untreated/T0 group set as the calibrator.

People attach entirely unnecessary value to specific methods, often to the point where they forget what the data actually is.

You don't need a calibrator. It adds nothing here.

Your questions are "Does miR-A go up with macrophage content" and "does miR-B go down with macrophage content"

You can measure miR-A and miR-B in (hopefully) multiple samples that you have designated "high macrophage" and "low macrophage". You can also measure U6 in those samples, and you assume U6 will not change with macrophage content.

Now you normalise miR expression to U6 expression (miR-A - U6, miR-B - U6) for EACH sample: these are your dCt values for miR-A and miR-B, in each sample, and represent the normalised expression of those miRs within those samples.

These are your numbers, though they are on an inverted log scale, so smaller numbers = higher expression.

Multiply all numbers by -1 to make it more intuitive.

Now these are your numbers on a log scale (which is where you want them, because qPCR data is normally distributed when left as log scale).

Are all the miR-A numbers in high macrophage samples bigger than those in the low macrophage samples? Is the reverse true for miR-B?

Those are your questions, and you now have the numbers to answer them.

What would a calibrator do here? The question you are asking is "is this set of numbers different from this other set of numbers", and that isn't a question that needs anything to be "calibrated to 1". The numbers themselves are entirely arbitrary, and if you used a different reference gene, might all be completely different, but the relationship between those numbers should stay the same.

-dCt values of 2, 3, 2 vs -1, -1, -2 : that's a significant difference.

If you add 5 to all the values such that you now have 7, 8, 7 vs 4, 4, 3, it's still a significant difference (it's exactly the same significance value, too).

It's the relationship between the numbers you are testing, not the numbers themselves.

Also note, the mean difference between the two sets IS your fold change, on a log2 scale. For the numbers above, it's 3.666 (in both cases), corresponding to a ~10-fold difference. I'm using -dCt values here, so the first set is higher expression: the first set has ~10x as much expression as the second set.

I don't think you need to explicitly describe this process at all: you have Ct values, you used your reference to obtain dCt values (normalised data), you then tested that. That's statistically valid and entirely appropriate here.

What MIGHT be useful is to include (in your manuscript) a ballpark figure for the raw Cq values, because a major issue with normalising qPCR data is that all abundance information is discarded.

If you had the data above, and presented that, showing a ~10-fold difference in expression, but also noted that "Cq values for the second set of samples were ~28-30", this tells the reader (and you) that what you're looking at here is low expression becoming effectively "no expression" (Cqs in the 30s are near the limit of detection, and are unlikely to be of biological relevance).

If instead you said "Cq values for the second set of samples were ~20-22", this represents MUCH more abundant expression, so here biological mechanism is "loads of expression" becoming "slightly less expression", with concomitant effects on how you interpret these changes.

In essence, the difference between the two sets of numbers is what you're finding out. The approximate Cq values of your GOI gives you rough idea of abundance. The combination of these two allows you to make biological inferences.

Hi John Hildyard ,

Omg..thank you so much for the through and very detailed explanation. You finally gave me the opportunity to sleep peacefully after many sleepless nights. Just comparing the delta Ct values of both groups was what i had thought too, but as there were no papers that presented that way, i became doubtful and wasn't sure on how to present the data confidently. I now have a very clear idea on how to proceed and present. Thank you very,very much for taking the time to write such detailed explanation. I really appreciate your help. Thank you once again, from the bottom of my heart sir.