Amplification

[pointilliste]

It's nice to be working from home. I have the radio on, a pot of tea, a Lindt bunny, and a cat. All are distracting from the qPCR results in front of me

Comments

[laevis]

Did you run them in triplicate? Then you'll know if the error is internal or external to the qpcr. In my experience, it is pointless to look for less than a 2-fold difference with qpcr. You can theoretically find it if you run enough samples, but remember that every cycle = a 2-fold difference, and I generally can't get less than half a cycle variation within a triplicate run of the exact same sample (using Roche 2.0 cycler, which is supposedly very accurate). You can also plot a standard curve using 1, 1:10, 1:100, and 1:1000 dilutions to determine the efficiency of the primers. You can get variation if you're normalizing to a house-keeping gene and not taking amplification efficiency into account.

[pointilliste]

Yeah, I ran in triplicate and the replicates are tight, only a quarter of a cycle difference between them! The difference (of 6 cycles!) actually is between the same brain region from two ostensibly identical rats. I ran multiplex and singleplex and the difference is present in both. However, the negative RT is coming up quite early so I think I might have contamination. But it's only coming up for 18S and not for either GOI, and only for that sample, so I'm not sure where the contamination is from. NTCs are clear.

You're right, running a standard curve is the next step, also because I want to use delta-delta Ct to do relative quantitation so I need to verify that the amplification efficiency of the two GOIs and the normalising 18S are the same.

Thanks for your comments.

[laevis]

Also, I used an ABI that used 96-well plates in my last lab, and the general rule of thumb I was taught was not to fill the plates more than half-way in one run, because by the time the reactions were ready, they would sit out too long. I have no idea if this is true or just the superstition of the laboratory, but it worked for me. Then again, I was measuring differences of over 100-fold, so it would have been really hard to screw it up.

[pointilliste]

Yeah, I only had 45 wells full for this run, and even that takes awhile, even with pre-made master mixes so filling halfway is not a bad idea.

Although sometimes you want everything on the same plate so you can compare as many samples as possible to the standards. I suppose if you put the plate on ice and cover it to minimise exposure of the probe to light then you minimise risk of sample/reagent degradation.

The AB product info suggests starting the run within 2 h of setup, and if you can't, put the plate in the fridge.