If you think you have enough DNA you could precipitate it and perform sequencing/diagnostic restriction digest/nested PCR specific for your recombined product. You could always try to isolate the band and redo your RPA, if its the correct template you should get more product that's more specific.

Having all those different melting temps usually means you've got a bunch of different products.

Hi Peters! Many thanks for your answer. In my experiment I used SYBR green I to identify the success of RPA based on Real-Time PCR instrument and graphs. As I knew that without DNA, SYBR green I only diffuses in solution and overspreads the week signal of Fluorescence. In attached picture, the Fluorescence intensity is really not sharp, Iam wondering the peaks presented in graph just are only the product of primer noise?

I examined this RPA result by a comparison with a normal Real-time PCR reaction using the same primers, DNA, only difference one point that I used Real-Time PCR master-mix contained all of dNTPs, PCR buffer,Taq and SYBR green I. As the result, I got the clear signals in both of amplification graph and melting temperature (Graph below, sharp peak presented a main product and low peak is dimer event). An expected band showed on 1.5% agarose gel and compared together with RPA product, they were the similar sizes. I have just been questioning what was happened in melting temperature of RPA, why it can not be identified clearly like a normal Real-Time PCR result?

Another thing is the enzyme collection in RPA provided in form of  dried powder, after assembling the composition in E-tube (primer, D.W, template DNA, Rehydration buffer) then mixed with dried  enzyme powder, the solution seemed to be sticky. May it be one of reasons which made SYBR green I highly concentrated inside the solution during RPA ? Thus, the peaks I have seen that are the mass of SYBR green I?

If you're using the same primers for RPA and PCR, cut out the band from the RPA gel, dilute it 1/1000000 so that there's no chance of any original template being there, then use 1ul of this with the PCR primers. If it's the correct amplicon it will amplify.

As Cole says, the melt curve bands are probably different products - just because primers work well in PCR reactions doesn't mean they'll also work well in RPA reactions (although they will typically work). There are also LOTS of DNA binding proteins in RPA reactions, maybe they are interfering with the melting behaviour?

Dear Matthew S Forrest!

Thanks for your extremely qualified suggestion and explaination. I will try this way. The reason you guessed related to DNA binding protein was also a good point to notice.

 Hi, Can I ask you what kind of product size is the best? about 200bp? I am doing RPA primer design now. THX

Shorter amplicons should amplify faster and the shorter the better.

Melt curve analysis will likely not work, especially since amplification is produced within a lower temperature range. Also, your graphs show temperature ranges from 60-95C. These temperatures will denature the polymerase and recombination in commercial RPA kits. Is your goal to find an optimal reaction temperature? If so, I don't think you need melt curve analysis, unless you are using it to determine if non-specific products are forming. If you do need it, I'd suggest editing the melt ramp/curve step to reflect RPA conditions (37-42) or even (35-45), post-reaction.