[BioHak]

If you believe it is an amp by-product, why not sequence a sample of it on a capillary instrument and verify?

If your goal is to remove the smaller fragment, we cleanup our samples with the Aggencourt AmPure product? We see significant dropoff of any products under ~90bps and nearly 100% recovery.

[captainentropy]

Unless there is something going on that I haven't figured out yet I think the peaks you are describing for the PE amplification are amplified primers as I am seeing them too. However, the size of the peaks are larger, around 125 bp, though there is a tiny peak around 66 bp as well. In previous libraries I've seen a very strong 150-160bp peak that I know was amplified adapters that resulted from me cutting too far down on the gel after the ligation step. I've already sent these current samples for sequencing, I'll know for sure in a week or two whether they are adapters or primers.

In the attached images are my last two libraries. One is a library made from 100ng of input DNA and the other I used 1000 ng. Other than that and the day I made them on they are identical in construction (also, I use a 1:30 dilution of adapter instead of 1:10). Another person in my lab is using the same paired-end protocol and he has seen the same pattern. It was also present in a no-template control reaction, so we're confident it's just amplified primers. He has been optimizing the PCR step and has found that raising the annealing temp didn't prevent the peak from appearing and neither did adding DMSO to the reaction. The only thing that worked for him was to optimize the primer concentration. The concentration of the primers Illumina provides (or tells you to use if you ask them) is 25 uM. Thus in the PCR the [final] = 1 uM. For standard PCR I've always used [final] = 0.4 uM, and that is for 25 or more cycles! Although I am using the PE adapters/primers I've been using 18 cycles of PCR as written for the single-end adapter protocol. I'm not sure if I used 12 cycles as written for the PE protocol that it would make much difference as the products amplified in the first few cycles are most important for the final outcome. I can always just run the PCR on a gel and purify the library from the primers but the problem is at the PCR step. I think the primer concentration Illumina is suggesting is too high. Any opinions on that?

[csoong]

The pre-PCR gel extracted product should be characterized in terms of OD ratio and concentration and this would act as a validation step prior to PCR. This pre-PCR quality may greatly influence the required amount of starting materials and number of thermal cycles required for PCR-ing. If we could share info about pre-PCR characterization, post PCR characterization, insert size, and PCR settings on successful sequence runs, this would be very helpful to the community.

[captainentropy]

agreed csoong. However, if you are using ChiPed DNA you might only have a few nanograms to work with. That is hard too quantitate much less get a 260/280 ratio.

In my lab we're pretty sure the peaks we are/were seeing come from the PCR primers. Once I get my sequence back I should know for sure though.

[JUdw]

....The concentration of the primers Illumina provides (or tells you to use if you ask them) is 25 uM. Thus in the PCR the [final] = 1 uM. For standard PCR I've always used [final] = 0.4 uM, and that is for 25 or more cycles!

1ul [25uM] in 50ul PCR reaction, [final] = 0.5uM