• I agree with the basic premise of your point
• However, I tend to initially screen my primers by RT PCR on agarose gels to establish the optimal annealing temp: That is the Temp at which your specific product is present at the highest level ( implying highest PCR efficiency) but with no extra non specific bands: This 'conflicting demand' is a compromise between the highest temp at which priming is specific and thus results in one specific product but not so high that the specific product is reduced in yield, implying too stringent conditions and thus lower priming efficiency: The former 'ideal' tends to approximate to Tm-2/3C in general and the latter at a higher temp
• Thus, for qPCR I will routinely perform a temp gradient in 1C interval from Tm-3C to Tm+2C
• I then import this actual optimised annealing temp into my qPCR cycling conditions
• I still however,  verify that I obtain a single peak by qPCR as sybr green qPCR is >?= 10x as sensitive as gel based RT PCR where is is difficult to discern products much below 10ng (whereas low copy number genes yielding Ct values of ~ 30 cyles) might be present at ~ 2ng and thus invisible on an agarose gel
• In other words, although gel based analysis can be useful for determining optimal annealing temp for low copy number genes even this application might not be possible, necessitating primer conc. and/or annealing temp optimisation by real time analysis
• I will add one final caveat: if selecting annealing temp based on agarose gel analysis be sure to amplify with Taq and not a high processive polymerase like phusion: This is because the AmpliTaq formulation used on most sybr green qPCR mix resembles standard Taq in terms of amplification efficiency versus melt temp much more than a high processive highly engineered new generation polymerase and thus signal to noise profile with Taq is much more congruous with behaviour in sybr green qPCR. In other words what you see on a gel with Taq translates much more faithfully into sybr green qPCR reaction kinetics than what you might discover with a gradient gel based on some thing like phusion 

The best way how to be sure about your PCR product is to do the Sanger sequencing of your amplicon, to compare your sequence with reference sequence. In the obtained sequence you can also see nonspecifity background when annealing temerature is not optimal.

As Marcela said, you need to ensure that you have PCR products, prior to the DNA sequencing and analysis of your sequences by comparing them with the reference sequences. That way, you will absolutely need a conventional method like agarose gel electrophoresis. Even with large scale PRC, it is still possible to run the gel.

Bodhisattwa, the melting curve verification is sufficient in my experience. Your real-time PCR analyses software should be sufficient in this respect. There should be no need to validate using an agarose gel under these circumstances.

It all depends on your QC of the qPCR design. If you have validated already (using conventional PCR) that your qPCR primers give a single product and that this corresponds to your predicted amplicon (e.g. by sequencing) than a melting curve analysis is sufficient. If not, than a melting curve will only tell you if there is one, or more, species of product in there. Not what it is.

• I agree with the basic premise of your point
• However, I tend to initially screen my primers by RT PCR on agarose gels to establish the optimal annealing temp: That is the Temp at which your specific product is present at the highest level ( implying highest PCR efficiency) but with no extra non specific bands: This 'conflicting demand' is a compromise between the highest temp at which priming is specific and thus results in one specific product but not so high that the specific product is reduced in yield, implying too stringent conditions and thus lower priming efficiency: The former 'ideal' tends to approximate to Tm-2/3C in general and the latter at a higher temp
• Thus, for qPCR I will routinely perform a temp gradient in 1C interval from Tm-3C to Tm+2C
• I then import this actual optimised annealing temp into my qPCR cycling conditions
• I still however,  verify that I obtain a single peak by qPCR as sybr green qPCR is >?= 10x as sensitive as gel based RT PCR where is is difficult to discern products much below 10ng (whereas low copy number genes yielding Ct values of ~ 30 cyles) might be present at ~ 2ng and thus invisible on an agarose gel
• In other words, although gel based analysis can be useful for determining optimal annealing temp for low copy number genes even this application might not be possible, necessitating primer conc. and/or annealing temp optimisation by real time analysis
• I will add one final caveat: if selecting annealing temp based on agarose gel analysis be sure to amplify with Taq and not a high processive polymerase like phusion: This is because the AmpliTaq formulation used on most sybr green qPCR mix resembles standard Taq in terms of amplification efficiency versus melt temp much more than a high processive highly engineered new generation polymerase and thus signal to noise profile with Taq is much more congruous with behaviour in sybr green qPCR. In other words what you see on a gel with Taq translates much more faithfully into sybr green qPCR reaction kinetics than what you might discover with a gradient gel based on some thing like phusion 

Once you have optimized your assay, then you don't need to run out your qPCR products on an agarose gel. I would run an agarose gel only during the optimizing and trouble-shooting stages of the project.

It is possible that unspecific amplified product may have the same annealing temperature, in case like this we cannot claim that melting curve is sufficient, as we do not know about the unspecific amplicon's characterization, to make sure that the product is the target, it may need to further process like (sequencing, gel electrophoresis)

in support to this , i would like to mentioned that in my project i got 2 bands on gel of 150 bp desired and 250bp undesired product sizes, however melt curve is very specific, single peak and Tm is 85.

anybody can explain the situation melt curve is good and gel electrophoresis two bands?