I noticed something similar using the TruSeq stranded mRNA kit. For every library made, we saw the expected size product ranging 220-260 bp and another major product 400-600 bp on agarose gels. Illumina tech support suggested the larger product was a "daisy-chain" or "bubble" structure as mentioned by Emma.  They also suggested denaturing gel electrophoresis to resolve this, but we never tried that. However, we simply used the absolute DNA concentration from Qubit/qPCR and converted to molar concentration assuming the average size of smaller product seen on agarose gels. These libraries all sequenced fine on a MiSeq at 12 pM loading. Using a customized library prep method where we added adapters using entirely different strategies, we never saw the "bubble" product. I suspect the bubble product is formed by over-amplification, which is difficult to control even when using the 10-15 PCR cycles suggested by Illumina.   

I had something similar to this with my cDNA libraries (NEB next ultra directional RNA  illumina kit). In my case the band was caused by a limiting concentration of the primers in  the enrichment PCR step (this creates a 'daisy chain' structure that runs at a high MW on the bioanlyser -but will not be removed by the beads).

To remove the bands from the libraries, I ran 1 extra cycle of enrichment PCR with additional primers etc.

Good luck,

Emma  

I wouldn't worry too much ... these things happen here and then but never actually caused problems (even if these are real fragments that would hybridise to the flow cell - which I doubt - the hybridisation will be highly inefficient with these large fragments and the bridge amplification won't work either). So I would just go on sequencing.

Best and good luck

Matthias

Our main concern is whether or not a peak is present. We often see trailing peaks, but the sequencing reactions are still successful. The only time we do not proceed to sequencing is if no peak is present.

If you have large amount of cDNA library, how about running an agarose gel. Bioanalyzer is too sensitive to some electronic signals, however agarose gel is not.

The lower signal of you cDNA library has a too small molecular size (about 300 bp according to your pdf file), is it what you expected? If not, the upper signal of your cDNA library might be the real cDNA. The large size of the upper signal might be due to insufficient RNA fragmentation.

I noticed something similar using the TruSeq stranded mRNA kit. For every library made, we saw the expected size product ranging 220-260 bp and another major product 400-600 bp on agarose gels. Illumina tech support suggested the larger product was a "daisy-chain" or "bubble" structure as mentioned by Emma.  They also suggested denaturing gel electrophoresis to resolve this, but we never tried that. However, we simply used the absolute DNA concentration from Qubit/qPCR and converted to molar concentration assuming the average size of smaller product seen on agarose gels. These libraries all sequenced fine on a MiSeq at 12 pM loading. Using a customized library prep method where we added adapters using entirely different strategies, we never saw the "bubble" product. I suspect the bubble product is formed by over-amplification, which is difficult to control even when using the 10-15 PCR cycles suggested by Illumina.   

https://dnatech.genomecenter.ucdavis.edu/faqs/my-libraries-show-peaks-larger-than-expected-can-i-still-sequence-these-pcr-bubbles/