Amplification curves jagged - (Apr/13/2012 )

My experiment was to set up an efficiency curve for a gene using MCF-7 breast cancer cell line sample cDNAs. However, my amplification curves do have the shape of a correct amplification curve (Ct values also seem correct) but are jagged or zig-zag throughout and even recognized as 'negative' despite neat and specific corresponding melting curves. This is actually a second repeat of the experiment and the first repeat yielded ideal results. Therefore, I am wondering if there is a possible mechanical problem with the machine (eg. alignment of the xenon lamp). The machine I am using is Roche LightCycler 480.

You did use your plate multiple times, didn't you?
That happens. No more than two runs of one plate, and even the second can be a little bit zig-zagged.
Try the High specifity setting in the 2nd Derivative analysis module, it should make correct Cts.