Barcoding PCR products - (Dec/11/2012 )

I need to barcode PCR products before sequencing. However, as my PCR is a multiplex, barcoding the primers are not an option. Is there a possibility to generate short DNA barcodes (4 - 6 nt) and use a template independant ligase to add it to the PCR product?

Terminal transferase will indiscriminately add bases to the end of DNA fragments, from whatever dNTPs are present, and will not stop, so sequences will grow quite long. I suppose you could code with four different bases, and perhaps with the twelve two-base combinations (if you can take chances of being wrong). I think you'll find that making this work reliably will be difficult and more trouble than it is worth.

Can you just make several singleplex out of the multiplex reaction? usually there's not much or no difference in the setup of the pcr.

Thanks a lot for the suggestions.
Converting back to singleplex would not help me. This PCR is for typing a bacteria. This particular multiplex can identify 23 types. If I convert back to singleplex, I need to do 23 individual PCRs for each sample. I am thinking of this method just to see if we can pool two or more samples together before additing the subsequent barcodes at the libarary prep step. I know it would generate a lot of complexity. But the library prepr kits are so expensive !

I don't understand why barcoding the primers is difficult. You can have 23 * 2 primers for each sample, with each of the collection of 46 having a unique tag. Yes, it is a lot of primers, but this may still be substantially cheaper than alternatives. Actually, you only need to barcode one of the two primers.

I am just conceptualizing this aactually. Say, if I add barcodes to an already equilibrated set of primers, I am worried about if it would affect the euilibrium and the thermodynamic properties of the multiplex.

If it is a few bases, I don't think it will have much effect.

You can, I realize, reduce the primer count a lot by making a set of forward primers coded with N different codes, and a set of reverse primers coded with M different codes. Using pairs in each sample reaction, you could have up to N*M samples with no duplicate codes.

Thanks a lot for the suggestions, very much appreciated. I will try to implement this and see if it goes well.

You should make sure you can sequence bases very near the end of the PCR product. Often there are issues with base calls at the very end of linear DNA fragments. You may want to insert extra bases 5' of the bar code in your primer.

Thanks again. The issue is more complex I think. We are trying to see the applicability of using NGS for amplicon sequencing with barcoding and multiplexing. So lot of issues to overcome. Will see how it goes.