No. Quantifying cDNA is pointless: the mix still contains unincorporated primers, residual RNA, and free nucleotides, all of which will absorb at 260.

Assume your cDNA synthesis is 1:1 (start with 1ug RNA, get 1ug cDNA).

You do not need much cDNA for good qPCR: I use 8ug as standard.

I would dilute your cDNA stock 1/10 and use 1-2ul of that just to check. Depending on target abundance you may be able to dilute even more, but don't unless you really need to.

1/10 dilution should be the bare minimum, however.

Do not use neat cDNA, always dilute cDNA.

@John Hildyard Thank you very much! But, I would like to clarify one more issue. What if I have (file in attachement) 44 samples (22 normal and 22 tumor) with different concentrations of RNA? Should I dilute all of them to certain amount? For instance, how much ng should I take for all samples for RT-PCR? And, if I take equal amount of RNA for Rt-PCR - I should not to dilute cDNA for subsequent qPCR, yes?

John Hildyard

Use the same _amount_ of RNA in each reaction, wherever possible. Don't dilute your RNA prior to the reaction, there's no point. Dilute it as part of the reaction (i.e. add X ul RNA + Y ul H2O).

A lot depends on the max vol your kit can take, but if I know that I can use say...8ul of RNA total, and most of my RNAs are 200ng.ul-1 or above, I'll use '1600ng' as my target. If any of my RNA is less than 200ng.ul-1 (so I can't physically add 1600ng without exceeding the 8ul cap), I'll just add 8ul.

Most kits can take anything up to about 2ug, so that's your upper limit. 1ug would be fine, for example.

Once you've made your cDNA, dilute all of it to the same extent (1/10 dilution for everything). Definitely dilute it, though.

You're not just diluting out the target molecule population, you're also diluting out the cDNA synthesis buffer components so they don't interfere with your PCR reaction.

John Hildyard thank you, but at the image attached to the question I have 9 microlitres of volume and 1-100 ng of cDNA quantity according to the manufacturer's protocol (qPCR). How can I dilute RNA as part of the reaction at RT-PCR stage? I have the lowest limit at pipette starting from 0,5 microlitres. So, from second image where concentrations of RNA are mentioned, I have such concentrations where I cannot make dilution in the range 1-100 ng at RT-PCR stage (for subsequent qPCR, considering 1:1 reverse transcription ratio) with aforementioned lowest limit at pipette. For instance, I have 432 ng/microliter and if I take 0.5 microliter and dilute it with water I have 216 ng of RNA in one tube anyway and that's too much. Wouldn't I need to dilute the RNA before adding it to the RT-PCR reaction mixture so that all samples in the RT-PCR and subsequent qPCR (1:1 conversion ratio) have the same concentration. I don't quite understand this moment

John Hildyard Or you mean to correct concentrations at RT-PCR stage to the same level diluting RNA as part of the reaction at RT-PCR stage (for instance , to 500 ng/microliter) and after RT-PCR dilute alll products with cDNA to the same extent (for instance, 1/10) to obtain concentrations of them in 1-100 ng range (in case with 500 ng after RT-PCR it will be 50 ng after 1/10 dilution of products for everyone)

Rakhimbek Bektayev

Maybe it would be easier to just walk you through how I would do it.

Isolate RNA.

Spec RNA: check the 260/230 ratios specifically, and for any samples with ratios

Ruslan Kalendar Хорошо, Руслан Николаевич, спасибо

John Hildyard Thank you very much! And one more question: "If you ever need to decide between a plate with "many genes for a few samples" or "many samples for only one or two genes" or even "all samples for a single gene", always, always do the last one" - why? Just interesting. Ok, I will do "all samples for a single gene" but, if I have normal and tumor samples what is preferred: normal samples on one plate and tumor samples on another? Or normal/tumor samples sequentially on one plate?

When you have two plates for the same PCR, you cannot be 100% confident that the behaviour of one exactly matches the other. If the humidity is slightly different that day, or the buffer used in the second plate is from a different batch, or slightly older, or the PCR machine is different (or older, or just having an off day), or the plates have tiny differences in well thickness/depth, then you might find that the two ostensibly identical PCR runs (same buffer, same primers, etc) have very slightly different efficiencies. It doesn't have to be much, but after 20+ cycles even tiny differences in efficiency can mount up.

So all the data from plate 1 cannot necessarily be combined with all the data from plate 2.

You'd need to include some calibrator samples (run 3-4 of the same samples on both plates) to allow you to adjust for variation between plates: this can be as much as an entire cycle, in my experience.

Alternatively, if you load all your samples (for a given PCR) on one plate, you know that all those samples experienced the exact same conditions, so all those samples are directly comparable with each other, for that gene of interest.

If you have half your samples on one plates, for two genes...you haven't gained anything really, because you're never directly comparing gene 1 with gene 2 (and you shouldn't: different PCRs are not directly comparable), but now you'll have to run two plates and run calibrators for both gene 1 and gene 2.

All qPCR is about differences between samples, not between genes, and loading all your samples onto one plates, for one gene, means you can assess those differences immediately, without need for calibrators.

EDIT: so yes, in your case, never do "tumour on one plate, healthy on another", that's pretty much the worst way to do it: you're not only introducing the potential for plate-to-plate variation, you're also ensuring that any variation that does occur will be exactly aligned with your different sample groups.

How many samples do you have? I run everything on 384 well plates, which lets me deal with very large panels of samples, but if you're stuck with 96s, you might need to be more cunning.

John Hildyard thanks a lot. So, what is the best allocation variant of samples on plate? I have N (normal) and T (tumor) samples. Both variants have 3 replicates. What is the best? variant1 (N and T samples on different plates) or variant2 (N and T samples sequentially)? Or there is no difference?

Variant 2 is better.

You would still need to include a subset of samples (say, two tumour, two normal) from plate 1 in plate 2 to allow for cross calibration, but variant 2 renders the bulk of your data immediately usable, and does not inherently assign potential random variation unevenly to your sample groups.