I am currently doing researching on a novel technique to detect antibiotic resistance genes by doing multiplex realtime PCR melting curve analysis.

It has been working so far, however, I am facing some problem recently running some samples, the problems are that the melting peaks shown in the samples are broad and unclear, the positive control does not show 3 peaks as desired and negative control show unspecific signal (A).

I tested the PC and NC again using new chemicals (B), thinking my working stock might be contaminated and my PC seems fine but NC isn't. I tried running with samples again (C) and the problems reappear. Please give me advice on what to do next or what might cause this?

Additional information: I am running with DNA samples, using SYBR green dye, there are also additional additives in the reaction (Glycerol and DMSO). The 3 melting peaks for PC normally appear at 76, 82.5 and 88 degree.