Protocol Online logo
Top : Forum Archives: : Real-Time PCR

TaqMan - 18S curve not sigmoidal, replicates are different - (Feb/21/2008 )

Hi,

I have run a 384-well plate and the results are useless.
First, the replicates are hugely different from one another, it is always the 2nd replicate that is gives a signal a few cycles (!) before the 1st one. The 2 replicates come from the same tube.
Second, the 18S amplification plot is far from being sigmoidal, it is quite flat.

Can anybody tell me what could have gone wrong?

Thanks a lot.

-Krisztina-

Hi Kristina,

Your replicates are obviously pipetting errors or not properly mixed. How do you do your replicates? Each well is prepared separately or one master mix for 2 wells? Its always better to do triplicates so you can delete the outlier and still have replicates. You should do one master mix for 3 wells, then take from this mix into the separate wells. Always mix everything properly. Non-sigmoidal curves indicate no amplification. What chemistry are you using? SYBR Green?

Cheers,
Chris

-chris_sylim02-

Hi Chris,

Thanks a lot for your reply. I think you were right and it was a technical problem due to improper mixing. As I figured it out, the centrifuge settings I have used to pull down the volume were changed by someone else and I ended up with a sedimentation of my mixture due to high-speed spinning. This has lead to this error always giving the 2nd replicate higher amounts.

Thanks a lot again, you helped me track this down and now is OK.
Krisztina

-Krisztina-

QUOTE (Krisztina @ Feb 21 2008, 10:30 PM)
Second, the 18S amplification plot is far from being sigmoidal, it is quite flat.


Could you post the 18S raw plot? 18S is commonly on very high expression level, too much template can do this, but can't tell you exactly without the image.

-Trof-

Thank you, guys, the problem has been solved. Unfortunately, my colleague has made a much higher dilution of primers and probes for 18S than we usually do and I seems likely this was the problem with the flat curve. Using the usual dilution gave me excellent curves and the experiment went technically well. Thanks a lot for all your kind help!

Could you post the 18S raw plot? 18S is commonly on very high expression level, too much template can do this, but can't tell you exactly without the image.
[/quote]

-Krisztina-