Having looked at both plots I would say that your signal to noise is low:
This accounts for your spurious traces but nevertheless at first glance the data still looks basically sound
Your amplification I also note starts at > 30 cycles suggesting that the copy number of your genes is low. This in part explains your low signal to noise
What do the traces from your housekeepers look like ? Do they give Ct values ~ 10-20 cycles
If these traces include your housekeeping primer amplifications then you have problem with low PCR efficincy for sure
That could be due to low quality cDNA preps which in turn might be the result of poor quality and/or quantity of RNA. How much RNA are you putting into your RT reactions and what is the 260/280 and 260/230 ratio of your RNA ?:
Your RNA should have a 260/280 ratio > 1.7 and ideally 1.8_2.1 and with that RNA you should use > 200ng in your RT reaction
I notice that your second dissociation peaks are are just one or two wells: If this corresponds to a single well for particualr gene(s) and the other duplicate/triplicate technical replicates are Ok then simply exclude them from your final analysis
Extra lower melting curve peaks can correspond to primer dimer (which relates to primer design) but unless you see this extra peak consistently with a particualr primer and if you have 2-3 other techncial replicates with a single melt curve for that primer I wouldn't worry too much (just exclude these outliers from the final analysis)
I would finish by saying that because occasionally you obtain spurious data from the odd well make sure you have an N of 3/4 for techncial replicates
I would also say that if you see additional spurious peaks in particualr wells as a consistent well related phenomenon I would clean that well or recalibrate your machine
If the effect is consistently related to a particualr primer pair then stop using those primer(s) and redesign, taking into account algorithms that deliberately screen for prier (homo or hetero) dimers:
To that end find a primer design SOP of mine which talks about how you do that
Actually I tested a primer prior to this run and it was all fine then I tested other primer which is this one for the same samples at the same day but this is what I got . I converted my mRNA into cDNA after storage of my isolated RNA in DEPC free water at -80c for less 2 days
I should note that I have diluted my samples (cDNA) by 10 folds
a) from the melting curves you seem to have a product (1 peak at one Tm and probaby with different starting concentrations) in a number of reactions, many are negative and some have by products
b) you see no amplicifations curves
At the first sight you seem to have a big baseline problem. As you can produce melting curves I think you use an intercalating dye.
Can you analyse the curves with a melt peak apart form the other reactions? Please do that and show me the curves without and with baseline setting and represented at a log-linear Y-axis.