It would help to see a picture of the amplification curves and have the actual ct values and dilutions.

Hi Jochen, can you open the .eds file if I send you?

No. Please use a more "common" format, e.g. text (or Excel) for numbers and png, tiff or jpg for pictures. Thanks.

I have attached the exported excel file and a picture of std curve. As you see in the std curve, if you connect the 1:2 dilutions, the slope will be different from when you connect 1:32 dilutions.

The image

Forget about the ct values for the lowest 4 quantities, they do not seem reliable. The other 8 points seem to indicate that the actual dilution factors were not what you think they are. I would guess that the 1:32 worked well but that the 1:2 is actually a 1:3 dilution series.

Can it be that you made teh 1:32 dilution as "1 part DNA plus 32 parts water" and the 1:3 dilution as "1 part DNA plus 2 parts water"? This would fit the data. But then you should calculate with a 1:33 dilution an a 1:3 dilution. Then the efficiencies are all around 1.74. The 95% CI for this case is from 1.70 to 1.79.

About the highest three Cts I agree with you. About the dilutions, as I always do 1:32, I pipet 2 times 15.5ul in each tube and take 1ul from the more concentrated, mix very well, and take 1ul to next tube (which has 31ul). Inaccuracies in this mode can be explained by for example a bad pipette, since I do not use the same pipette for 1ul and 15.5ul. But about the 1:2 dilutions, I use the same pipette, so even if the pipette is wrong the dilution should be 1:2. I had 10ul water in each tube and took 10ul from the previous, mixed well, and took 10ul to the next. I will try just a few 1:2 dilutions again, to see if I have had by mistake 20ul water in tubes. I hope this is the case ;)

But still the question is why I get such low efficiencies using the same primers, stds, and SYBR green that I used months ago! this is quite annoying, as I see a general decrease of efficiency in all my qPCRs (different primers) recently.

Thanks for your help. Yes, degradation could have happened to some extent, but I think this would not influence the efficiency calculations, since everything together (intact and degraded DNA) is diluted equally, so overall shift of Cts will occur, but the distance between Cts, i.e. the number of cycles the reaction should work to make a certain number of DNA strands 32 times more, should stay the same. I use water for dilutions, since I have water in my samples, so having TE in stds may change the efficiency of reaction only in stds, which will influence all calculations on the samples.

Maybe it is worth mentioning again, that I make the dilutions always fresh, the only thing which is kept in -20 is the concentrated aliquot, which comes up at Ct of 10.

During the creation of your dilutions you may want to consider weighing water before every pipet delivery to prove the volume it is delivering. E.g., 10 uL of water at RTP should weigh 0.01 grams on an analytical scale. I do this with all qPCR with all pipets used to avoid problems like this. If your dilemma is only pipet-associated, this could help. Avoid using DEPC water to store your cDNA - it can cause degradation of your sample. TE pH 8.0 is a much better bet. Dilutions can be performed in nuclease-free non-DEPC-treated water. Creating a dilution scheme wherein 1 or 2 pipets do not have to be adjusted at all in between serial dilutions is always advisable - but, again, weighing water beforehand can help you avoid errors there as well. To deliver 10 uL for example, you may have to adjust your pipet to 9.3 uL or 11.2 uL -- even though it reads wrong, if the volume shown on the scale says 0.01 g, then, that is the true delivery...

@Jack: "that is the true delivery..." - given the scale is calibrated better than the pipet ;)

PS: Have you ever considered doing this in a 4°C cold room? Or do you adjust for the density at the room temperature? (e.g. take the scale result x 1.002 at 20°C)

PPS: just kidding

PPPS: one must judge if the achieved accuracy is sufficient, or - looking from the opposite direction: if the errors are of a size that the can be made responsible for the problems one observes.

How true - the scale could be misleading too... Thanks for the 20°C density adjustment factor - that is a crucial missing piece! ;]

Let me complete this nice talk by mentioning the last missing piece: also the gravity depends on the altitude of where you are ;)

I miswrote a few entries on this site then - not taking the proper density factor into account regarding the temperature at which the weight of water is measured and then converted into a volume.

So the factors would be:

factor adjust (mL/g)

4°C 1.000025001

20°C 1.00179612

21°C 1.002008526

22°C 1.002231468

23°C 1.002464559

24°C 1.002707812

25°C 1.002960841

26°C 1.003223659

So if water (on an impossibly good/trust-worthy scale) weighed 0.009975415 g at 23°C, the pipet would be delivering a volume of exactly 10 uL. (E.g., 0.009975415 x 1.002464559 = 0.01 mL = 10 uL). There we have it~!

Now I can finally move my scale out of the cold room ;]

it's been a long cold 8-year winter.

Amir - I think I will move my scale underground then - where only neutrinos would occasionally interrupt my operation ;] Ah - the gravity of it all.

Jochen always seems to provide food for thought on/in this forum - I feel like the assistant in National Treasure (with Nicholas Cage) - when after a brief exchange about daylight savings time (wherein this time, the assistant solved the riddle) - the assistant exclaims something to the effect: "So this is how you must feel all the time!"... ;]

Another thought does come out of this mini-topic though... and that is the idea of pipet calibration. Is it usual practice that when pipets are calibrated, let's say at 23°C...

Does the pipet calibrator (when weighing water) wait to see the number "997.5415 g" or at least "997.542" before he/she realizes perfection (1000 uL or 1 mL) has been obtained by the pipet at 23°C? It may be a small imperfection - but, perhaps with large scale PCR/qPCR... this could amount to a credible influence on mastermix assemblies, etc...

More food for thought. Thank you Jochen~!

SYBR Green mastermix can age over time -- are you using new mastermix too?

Yes, we consume lots of it, and no batch survives more than a week :) and expire dates are next year.

Is your 'stock' cDNA that has been stored, or gDNA of some kind? Human? Mouse? plant?

you mean the standard stock? it is PCR product, 1:5000 diluted, which makes Ct~10.

Do you use Roche SYBR Green with 'Low ROX' (30-50nM final ROX) added and take advantage of its corrective for Cq analysis? Or are you not using ROX at all?

This is what we use:

http://lifescience.roche.com/shop/en/us/products/faststart-universal-sybr-master-rox-

It has ROX.