Thanks for your reply. Really appreciate it!

I tried changing the baseline settings and it made no difference. It's set up for auto Ct, but I checked through each reaction to see if it had called it correctly. Might have another play though.....

Are these hump shapes common in all your wells or only to select wells. It is likely that you are amplifying two products instead of one. If the hump is ahead of your normal PCR product, it may be primer dimer. The easiest way to check would be to run your PCR product on an agarose gel ( I am sure you have discarded all your previous efforts) and observe if you have 2 products. Only a suggestion - this is seen in SYBR based reactions, not in probe based where strictly your probe should be sticking only to your amplified PCR product.

I guess you are talking about the melting curve (or is it the amplification curve?). If melting one it looks like a primer polymerization, one option is to rise a little bit your melting-amplification temperatures (depending on the your cycling settings) or sometimes adjusting your salt concentration also works. Of course there is the option of unspecific amplification or isoforms (that could be only a few bp different but stills detected by qPCR machine), to check that, once you see if you have only one band (and of the expected size) you could run a new gel, cut the band, purify and send for sequencing...

Hi everyone, thanks for your replies and advice!

I'm using Taqman probes. I have a generated a standard curve out of cDNA of known concentration, and the Ct values for the housekeeping gene range from around 14 for the highest to 33 for the lowest. For the gene of interest, the highest starts at 35 and the lowest is undetermined since I'm running 50cycles and the reaction hasn't started at that point. The hump shapes are only found in the wells with sample in, not the NTC or other wells. It is the amplification curves that show this, not melting curves. I will try running the products on a gel to see what I have!

Hi everyone,

I ran the products on a gel and the bands were faint compared to the ladder. There looked like there might be two products though, one at less than 100bp. Could be primer dimers?

If they are primer dimers you should have a band of the same size in the NTC lane, if not it could be an unspecific product or an isoform of your interest gene. In that case if you send to sequencethe band of expected size and the other product you could know what is happening in your PCR. By the way how are your melting curve, Have you a single peak?

I do have a band of the same size in my NTC lane!

Might not be able to sequence as my supervisor is short of money. I've not had a chance to do the melting curves, sorry.

Thanks for your input!

Good for you, it looks like they are primers then.

About sequencing you could think later in that (it costs about 8 euros by sample, so is not really so expensive). And for melting curves, I would suggest to add it in your settings for next qPCRs, normally you just add that at the end of your amplifycation cycling and the machine should analyze that and give you at the end the melting cuves as well as the amplification curves, the melting curve is very important to assess the quality of your analyses by qPCR.

Good luck with your experiments!

Ok, I'll try that next time. Thanks!

If the humped shapes aren't primer dimers, any other ideas what they could be caused by? I've tried reducing (1 in 10 dilution) and increasing (double volume) the amount of cDNA, increasing the number of cycles (from 40 to 50) and increasing the annealing temp (from 60 to 65). Should I try changing the amount of primer? Or reducing the annealing temp?

Your cDNA is good because your HKG works. Next -I would check whether your sample is actually expressing your gene of interest. You mentioned that the Ct values for your highest conc. came up at ~Ct35 - which means that there is hardly any expression in your sample. Suggestion - You can try out a new primer set that amplifies a bigger sized PCR product and run a regular RT-PCR on a agarose gel to verify whether your gene of interest is actually present in your sample. Use varying amounts of cDNA. No point in running your TAQMAN, if it is not there.

Thanks for your input.

Unfortunately my supervisor is not keen on getting more primers as he is short of money :(

Have you tried doing the standard curve for your samples using at least 5 point dilution? My experience is that there shouldn't be any bubble in the tube to get a R^2 of 0.99. If you are using plates, that is also recommended. I tried running my qPCR ten times and found out in the end that vortexing followed by mini centrifugation for at least few seconds really helped in eliminating bubbles. I don't use probe, I only use sybrgreen supermix at RT, same primers at 3 different primer dilutions for comparison and same annealing temp I used for standard PCR.

Thanks for the advice.

I haven't tried doing a serial dilution for my samples, mainly for two reasons - firstly my actual standard curve is based on a sample of known concentration, and secondly because I am using only 20ng sample in each reaction, which is quite near the limit as far as I understand. I can't use too much as the samples are precious.

I am currently using plates, and I vortex the mix before putting into the wells. Once the samples have been added I pipette mix the wells then spin the plate briefly.

I'll try different primer dilutions, as you do.

Thanks again!

Hi everyone,

So, I've tried doing my reactions in singleplex (i.e. without the housekeeping gene) in case the primer for the housekeeping gene was outcompeting the other primer, and I've tried the primer at double volume and at a 1in10 dilution (again without the housekeeping gene in the reaction). None of these made any difference. I might try a further increase in annealing temp and/or adding DMSO. I'm increasingly thinking that the primer is rubbish, and fortunately so is my supervisor, so I might be able to convince him to get a new primer, or at least do a SYBRgreen assay to check efficiency :)

Is it OK to ask what is your gene of interest and what kind of tissue lysate you are working with? Perhaps somebody has a tried and tested pair of primers that you could use to analyze your samples.

Hi, Anne, you need a positive sample which express your gene of interest and test if your primers work or not by common PCR before moving to realtime PCR.

Thanks Wilco!

Anuradha, my gene of interest is SULTIEI, the gene for Estrogen Sulphotransferase. I have lysates of colorectal cancer cell lines (COLO205, HCT116, HT29, also JEG3 as a positive control).

Thanks for your interest!

I think you should try out a new primer set.

Thanks for your input, Alexander. Hopefully I can convince my supervisor!

Hi everyone,

Bad news, we bought a different probe and the results are still the same! Any more ideas? I've also tried using a different mastermix, but no success :(

Hi Anne, I think is better if you make your tests with classical primers instead of taqman probes. I mean the typical 18-23 nucleotides oligos that you can buy in several enterprises and design by yourself based in the sequences available, thus you can manage more parameters such as technical conditions but specially the location of the fragment your are amplifying which would allows you to differentiate among alternative splicing and/or isoforms. In addition you could sent to sequence the fragments you obtain. All of that should be for the same or even lower price than buying a taqman probe. Once you find the best ones you could decide if going with taqman probes or oligos for the qPCR.