20 replies to this topic

[molecular_medicine]

Hi all,

I am a newbie to qPCR.. I did a couple of runs of qPCR. I used the power SYBR green kit from applied biosystems, and used pre designed primers. I did have a bit of contamination in a few samples (checked it with my RT- samples). I am sorting out other issues. For now, please have a look at my dissociation curve, people find it very weird but none has an idea why this would be happen. Can anybody please help me trouble shoot this problem!!??!! pleeeeeeease!!!!!!!!!

Attached File(s)

•  melting_curve___norm_hypo_2_hrs_mda.JPG (129.97K)
  Number of downloads: 549

[pcrman]

The dissociation curve does look weird. Have you run a regular PCR with the primers and your samples? That may give you an idea what is wrong.

[molecular_medicine]

Hi all,

Sorry I figured out why it looked very weird.... I did not omit the empty well during analysis... now that I did it, my melting curve still looks weird, but better than before. can you please pour ur suggestions as to why it still looks weird?? please!!!!!!!!! thanks...

Attached File(s)

•  melting_curve___norm_hypo_4hrs_mda.JPG (209.42K)
  Number of downloads: 357

[littleaxt]

A bit overloaded with information, your little picture. Maybe you can toggle of a few samples. Anyway it looks like you have a lot of background fluorescence and primer dimers.

[aimikins]

I agree. have you done the titration curves to determine reaction efficiency?

[Curtis]

I never trusted plate based real-time from the beginning. I would still go for the rotor based machines like Light Cycler 2 or RotorGene 6000

[aimikins]

I've had quite a bit of luck with it in the past, but getting your conditions squared away in the beginning can be a real nightmare.

[molecular_medicine]

Curtis, on Feb 26 2009, 09:40 AM, said:

I never trusted plate based real-time from the beginning. I would still go for the rotor based machines like Light Cycler 2 or RotorGene 6000

I know, light cyler or rotorgene is really good and effective.. but our very poor labs have just this plate based systems....!!

[molecular_medicine]

aimikins, on Feb 25 2009, 10:21 AM, said:

I agree. have you done the titration curves to determine reaction efficiency?

titration curves?? can u please elaborate on how to do that or give me a link which will help me understand the procedure better.... thanks....

[molecular_medicine]

Thanks for the response guys...

For now, I am testing my primers on human genomic DNA, hopefully the bands r on the right spot and please god please, no primer dimers!!!!!!!!!!!!!!

PS: I got my primers from the literature, they used these primers in taqman, but I just got the sequences and am using it for SYBR green, did a BLAST search and it returned with just my gene of interest.. Should this be ok???????

[aimikins]

qPCR is very, very sensitive to the concentration of primers, even if no dimers are anticipated. if you can, find ABI's User Bulletin #2 on their website. it's a pain to read, but it tells you how to do titrations of primer and template in order to get a good reaction efficiency. if you don't do this, you're not likely to get any usable results.

[gfischer]

molecular_medicine, on Feb 24 2009, 08:44 AM, said:

Hi all,

Sorry I figured out why it looked very weird.... I did not omit the empty well during analysis... now that I did it, my melting curve still looks weird, but better than before. can you please pour ur suggestions as to why it still looks weird?? please!!!!!!!!! thanks...

It's hard to tell what's what with this many lines on one graph, but it looks like each line has only one major peak, which is what you're going for. However, if all the wells are the same gene, you still have a problem, since there is a wide range of Tm's.

[josurb]

molecular_medicine, on Feb 24 2009, 08:44 AM, said:

Hi all,

Sorry I figured out why it looked very weird.... I did not omit the empty well during analysis... now that I did it, my melting curve still looks weird, but better than before. can you please pour ur suggestions as to why it still looks weird?? please!!!!!!!!! thanks...

Ok, what I see when I look at this, is that you're looking at the melt curve for multiple samples right? If so, what you're seeing is all of the different melting curves, rather than an individual sample's. You need to select the wells that are for a particular primer set and look at them. Then you should see just one peak hopefully. Some of it still looks like there may be primer dimer in some samples or what not, but you have to look at the samples individually to make sense of it. Ex. (Look at one tissue, one primer set, all wells with this combo). Hope this helps.

[beth]

I would suggest just to run the PCR machine with either no plate or a new empty plate. This will let you see if the machine is contaminated. I visited a lab who had funny plots like you. They ran it without a plate and got all sorts of amplification, so the machine had been contaminated by someone. They cleaned it and then everything was fine again.

[UBClabbie]

I used plate real-time pcr (Roche 480 --> 384 wells) and I have found that there are many conditions that can affect the specificity and efficiency of your amplification. You may need to tinker around with your experimental setup in order to optimize it. I suggest running a few test runs with different qPCR programs (ie. 2 step vs. 3 step, different annealing temperature) and also using different amounts of starting RNA. Different machines behave differently, so unless your supplier tested the primers and mastermix for the specific machine you are using, its unlikely that you will get perfect results.

If your dissociation curves aren't looking so good its probably caused by 1 or both of these reasons:
1. your sample is contaminated with proteins and other junk or is degraded, ie. your sample has low quality.
2. your primers are not specific enough and are amplifying other targets - redesign your primers or adjust your mastermix formulation, increase annealing temperature to increase stringency.

Edited by biotechgirl, 16 December 2010 - 02:14 PM.