I've had some experience with both Illumina library preps and Ion Torrent library preps and I am thinking about efficiencies on a basic molecular level. Ion Torrent/454 preps use blunt end ligation of 2 different adapters, while Illumina uses TA-ligation of 1 adapter. Correct the following logic if I'm wrong, if we have 4 dsDNA molecules:
Illumina's protocol should arrive at 8 effective ssDNA molecules, for 100% efficiency. The Y-adapter prevents misdirectional adapter ligation, but allows adapter-adapter ligation dimers to still form.
Ion Torrent's protocol should arrive at 2 effective dsDNA molecules, for 50% efficiency, as 25% of the fragments will have A-fragment-A and 25% will have P1-fragment-P1. Granted, these will be non-amplifiable but they do sequester true DNA fragments in solution. Additionally, the blunt-end ligation protocol would hypothetically also cause a whole mess of chimeric ligation products, such as A-fragment-fragment-P1. Fragments shouldn't ligate on both sides of an adapter due to the double T tailing, which also prevents misdirectional adapter ligation. Adapter ligation dimers will still form, at much the same rate as Illumina adapters (?), though composed of 3 combinations of the 2 adapters.
Any issues with this? Why didn't Life Tech go down the TA route as that would solve a lot of the blunt-end problems? And yes, I am aware that there are some groups that have developed their own Y adapters for use on the Ion Torrent/454.
Looking forward to comments and/or kind remarks that someone else has already talked about this
Illumina's protocol should arrive at 8 effective ssDNA molecules, for 100% efficiency. The Y-adapter prevents misdirectional adapter ligation, but allows adapter-adapter ligation dimers to still form.
Ion Torrent's protocol should arrive at 2 effective dsDNA molecules, for 50% efficiency, as 25% of the fragments will have A-fragment-A and 25% will have P1-fragment-P1. Granted, these will be non-amplifiable but they do sequester true DNA fragments in solution. Additionally, the blunt-end ligation protocol would hypothetically also cause a whole mess of chimeric ligation products, such as A-fragment-fragment-P1. Fragments shouldn't ligate on both sides of an adapter due to the double T tailing, which also prevents misdirectional adapter ligation. Adapter ligation dimers will still form, at much the same rate as Illumina adapters (?), though composed of 3 combinations of the 2 adapters.
Any issues with this? Why didn't Life Tech go down the TA route as that would solve a lot of the blunt-end problems? And yes, I am aware that there are some groups that have developed their own Y adapters for use on the Ion Torrent/454.
Looking forward to comments and/or kind remarks that someone else has already talked about this
Comment