[mollusc]

I again feel it's possible that SS libraries would form hairpins of various kinds that would bind dye but migrate aberrantly.[/QUOTE]


Thanks for your helpCnicolet!

I always had this question.What are the factors controlling the formation of secondary structures of ssDNA.Is it just the repetitive sequences or anything else as well?

[sulfobus]

2nd larger DNA band after enrichment PCR

I frequently see this phenomenon, not only realted to Illumina sequencing. I believe it is mostly heteroduplexes, i.e. partially ssDNA that partially hybridize to other strands with a segment of similar sequence. This will form bulges of ssDNA that will slow gel migration, and can explain how it could appear just by denaturing/renaturing the DNA.

It is also seen in Quail et al, Nature Methods, 5:12 (2008). As stated in that article and in this and other threads, limiting the number of PCR cycles or limit the DNA load will decrease the formation. It is formed during the plateau phase of the PCR.

Other PCR artefacts take place in the PCR plataue, most notably recombinations, so there are many reasons not to dwell in the plateau for too long.

[tomv]

We have an interesting twist on this issue. We have had samples that have been gel purified post-PCR and then, with nothing else happening to the samples, when we run them on a HS-BioA chip we see TWO peaks with the second peak ~2x the size of the gel fragment. This would seem to rule out the asymetric PCR theory, as those ssDNA products should have been removed in the gel. Could it be that the dsPCR products are "recombineing" at the adaptor sequences to form heterodiamers?

[andibody]

When we saw those double bands in summer last year we started to have additional problems, especially a reduced efficacy of adapter ligation and a decrease of unique reads. We started using NEB sample prep reagents, using only paired end adapters (also for single read runs), aliquoting dATPs in minimal amounts, doing a-ligation and adapter-ligation very carefully (never stopping in between these two, vials always on ice, etc.) and never had again any of mentioned problems including double bands. What exactly resolved the problems, I can't tell, not even if the problems we saw were directly related to one another but I suppose the solution was one of that modifications to the protocol. This is a highly anecdotal story, sorry, but I thought it might still help someone.
protocol: 5ng starting material, adaptors 1:20, pre-PCR gel purificacion,18 cycles PCR.

[HCMVSeq]

Same problems with PCR enrichment

Hello Everyone,

I've had the exact same problem with my Illumina PCR enrichment step prior to sequencing...I've had two different-sized PCR products show up...one that is my expected size (about 250-300bp) and another that is about double that (500-600bp). I've done a PCR "cycle titration" to see when each band shows up during PCR. The attached image "PCR Illumina 3" shows my initial PCR product after 17 cycles. The attached image "PCR Illumina 2" shows a cycle titration every 5 cycles from 10 to 35. As you can see, cycle 15 shows the lower (expected product), but the remaining cycles show just the larger product. Finally, "PCR Illumina" image shows a finer cycle titration (every cycle from 11 through 18 except 14...forgot to take sample ...) and you can clearly see a shift from the lower MW product to the higher MW product. My solution is to do a gel extraction of the lower MW band at cycle 15 (most lower MW product...most likely still in the exponential phase of PCR)...I chose to do a gel extraction to remove any primers and possible "primer/adapter concatemers" that have been discussed in this thread and others (didn't see any on gel, but I guess it's better to be safe than sorry). I'm not sure what this shift in MW means...but I like the idea of ssDNA secondary structures that has been mentioned in this thread.

Good luck everyone,
Pete

[AdamB]

HCMVSeq: Thanks for the informative post. Please can you tell me how much input DNA you used for the library prep? And did you use the Illumina primers or your own?

[Nik]

[B]2nd larger DNA band after enrichment PCR[/B

I've seen this and I diluted my sample down and it went away (higher concentration samples). I've also (if the library is strong) cut it out of the gel again and that works a treat as well. The second band does make working out the concentration hard and as far as I've seen does have an impact on cluster numbers if you ignore it.

[KAP]

Hi everyone,

I have a variation on this problem which has me completely stumped - I would REALLY appreciate any ideas/help!

- I am seeing the a double-band phenomenon in my PCR gels, with a correct band at ~ 220 bp and the larger one at ~300 bp.
- I cut out the correct product, gel purify it and run it on the bioanalyzer. The same two bands appear.
- So I run it on a gel again - very far this time! - to completely separate the fragments, and gel extract the band again. This time I am SURE there was no accidental excision of the upper band.
- Run it on the Bioanalyzer again -> the same two peaks, at EXACTLY THE SAME RATIO.

So, I conclude that whatever it is, it is reforming in my sample, after gel extraction??

This would all not be a problem, except that when these libraries sequenced, the cluster number was very low (1/3 of what it should have been) - so I assume the high molecular weight band does not form clusters.

Please help - any ideas at all!!!

Exp details:
- Home-made adapters used (with the recommended modifications).
- PCR - x10 cycles, with quite a bit of DNA (probably ~25 ng in a 50 ul PCR).
- On someone's recommendation, I switched from Paired-End to Single-End adapters but no difference.

[GW_OK]

Do you suppose it could be something physical like DNA breathing, supercoiling (doubtful...), or strand interaction?

[KAP]

Do you suppose it could be something physical like DNA breathing, supercoiling (doubtful...), or strand interaction?

Thanks for your (quick!) response! DNA breathing? I've never heard of that! It's cDNA (for RNA-seq), does that change anything? Strand interaction is a possibility... I'll look into that. Thanks again!

[csquared]

Many, but possibly not all, of the above situations of higher molecular weight bands on the Agilent Bioanalyzer can be caused when primer becomes limiting during the PCR. This can be caused by too many cycles, too much DNA input or too little primer.

A simple test is to take the library that you observe the two peaks in, add PCR reagents and perform 2 cycles of PCR. If the peaks collapse back to the expected size, you were seeing bubble products in the original sample. The bubble products will sequence just fine so if you see the two peaks, a real-time PCR quantitation of the library will be most accurate for establishing input into clustering.