noisy/spiky peaks on bioanalyzer

[riehle]

Hello!
I am kind of new in this NGS-business, struggling with library prep.

I prepared Testlibraries with ChIP'ed DNA using the TruSeq DNA Sample Prep Kit. I know that this kit is optimized for an DNA input of 1 µg, still I started with 10 ng.
Several changes I made to the original protocol including:
MinElute based purification
50-fold diluted Adapters
18-cycles of PCR
Furthermore I did the PCR prior to gel-based size selection.
From the gel I cut: 300-400 bp fractions (samples 4-6) and 200-300 bp fractions (samples 7-9 on bioanalyzer file) and purified the DNA with the MinElute Gel extraction Kit.
The library was validated using the High sensitivity DNA Kit on the Bioanalyzer.
I attached the bioanalyzer file. Please have a look at the samples 4-9.
Why are these peaks to spiky and why do I have another smear band higher than the desired fraction....
Advise is more than welcome...
Thanxx

[TonyBrooks]

The smear is a result of over PCR amplifcation. Drop the number of cycles.
Also, the general noise on the Chip is probably because you've over-loaded it. We found the HS chip does strange things when you load more than 10ng. Estimated yield from sample 2 is >100ng. Overloaded wells also interfere with subsequent samples.
I would dilute everything 1:20 and re-run the chip.
Not sure about the spikes though. Did you follow Ethan's protocol?

http://ethanomics.files.wordpress.com/2012/03/chip_truseq1.pdf

[riehle]

Well thanxx TonyBrooks
I did not use Ethan's protocol.
I followed mainly the TruSeq Protocol and used their reagents (see attachment)

Even if I do overamplification of my samples resulting in this second smear band. How comes that i can not get rid of this much higher molecular weight band by gel out after PCR. Somehow this second smear band must be a result of the desired fraction???

[TonyBrooks]

So you gel-cut after PCR?
What happens when you over PCR is that the primers begin to run out. When this happens the adapters at the end of your fragments can anneal to each other (rather than to primers) leaving a bubble in the middle where there is non complimentary sequence (it's highly unlikely that once denatured, the library will anneal back to it's compliment). These bubble fragments probably migrate at a normal speed through an Agarose gel, (after all they are still the correct bp size) but they will run slower on a Bioanalyser as bubbles and single-stranded DNA are an issue.
It's not a major problem as long as you can quantify your library properly. They will still form clusters with the correct insert size on the flowcell. However, it's better to reduce the PCR cycles as your library looks cleaner on the Bioanalyser and you should get lower duplication from your sequencing.

[riehle]

Yes I gel-cut after PCR.
I have never heard of bubble fragments. So just to get this right:
If I run out of primers, the adapter ligated DNA-fragments start to hybridize to each other resulting in this bubble-fragments. These bubble fragments migrate on the Bioanalyzer at a higher molecular weight compared to the fully complimentary fragments?
Do you think reducing from 18 to 16 PCR cycles might already have an effect on to this or should I reduce to 10 or so PCR cycles.
Thank you very much.

So you gel-cut after PCR?
What happens when you over PCR is that the primers begin to run out. When this happens the adapters at the end of your fragments can anneal to each other (rather than to primers) leaving a bubble in the middle where there is non complimentary sequence (it's highly unlikely that once denatured, the library will anneal back to it's compliment). These bubble fragments probably migrate at a normal speed through an Agarose gel, (after all they are still the correct bp size) but they will run slower on a Bioanalyser as bubbles and single-stranded DNA are an issue.
It's not a major problem as long as you can quantify your library properly. They will still form clusters with the correct insert size on the flowcell. However, it's better to reduce the PCR cycles as your library looks cleaner on the Bioanalyser and you should get lower duplication from your sequencing.

[TonyBrooks]

There are load of threads here on that subject as it's a common occurrence.

http://seqanswers.com/forums/showthread.php?t=20277

http://seqanswers.com/forums/showthread.php?t=20703

http://seqanswers.com/forums/showthread.php?t=17518

There's also a link somewhere in there from Ethan again about how to optimise your PCR.