Truseq RNA library has 2 peak on bioanalyzer

[houda]

Dear all,

I've been making Illumina truseq cDNA library and I've noticed 2 peak on bioanalyzer (DNA 1000) after PCR amplification. Is there anyone seeing this kind of result before? The second peak is > 1500pb!!!

What can I do to avoid this problem?
Thanks!
Houda

[niceday]

what does your peak look like before PCR and did you cleanup using ampure? I have seen some extra peaks if my libraries are not completely clean of ampure beads. Usually at the higher end of the spectrum.

if you are worried size select.

[dhinerfeld]

We have seen this. If you go into the PCR on the high end of target amounts, in the late cycles of PCR the primers become limiting. At this point, when you anneal, you get many bubble molecules where the two adaptors on either end of the fragment anneal to each other but the fragment in between does not anneal. These bubbles run larger.
We saw these larger species in the bioanalyzer runs but got great sequencing data. When we automated our prep on the Caliper robot, they went away. I think the reason for this is that the yield of adaptor ligated fragments is less on the robot. We know that in general our yields on the robot are lower than when we did it manually, although sufficient.
I have never done it, but someone suggested adding more primer and doing a few rounds of PCR and the larger species/bubble will disappear.
I'm not sure about all of this, but it makes sense to me.

[bruce01]

Yes, we have this also, Dhinerfelds answer is what Illumina techsupport told me: the template forms 'concatamers' due to the adapters annealing.

We have thought about reducing PCR cycles, increasing PCR reagents or reducing template. I have been looking at DNA quantity post dscDNA and end repair stages. Post dscDNA~6ng; ER ~25ng/ul (on nanodrop); gives the 'bump'. I have not tried anything to solve the 'bump' and Illumina don't seem too worried. As they do not have adapters at either end they will not sequence (though universal adapters could hybridise and lower clustering...)

[sblake]

Yep, the double peak. Very familiar and common to several users I spoken with. The primary peak is usually at ~225 and another around 500bp. Specifically relating to the TruSeq RNA kits, I wonder if the observation isn't also due to lack of complete chemical shearing as well as the bubble molecule idea. Library always sequenced well.

[upendra_35]

Hi

I did see double bands when i run my TruSeq libraries on the Agarose gel but haven't seen it when i run the same samples on the bioanalyzer. So i guess it is something to do with my running on the gel. However before running on the bioanalyzer i did contact the Illumina Tech guys regarding the double bands and this is what they have to say...

"We see customers' sequencing libraries showing a higher second peak from time to time based on Bioanalzyer results or double bands in a gel. The second peak may represent single-stranded library products that have self-annealed. This may occur because too much starting material was used at the beginning of library preparation or because too much template was used in the PCR reaction step. (NanoDrop and BioAnalyzer results do not give accurate quantitative results. We recommend Pico Green or qPCR).

As long as the second peak represents only a small percentage of your sample, you can proceed with cluster generation using these samples. However, if the second peak represents a significant percentage of the library, it may affect accurate quantitation of the sample. In this case, you may want to quantitate the sample with a method that detects only double-stranded DNA (pico green, qPCR). You may also re-purify the library by running a gel and cutting out the band of interest".

I hope this helps to explain your two peaks.........

[sblake]

post deleted due to finding good explanations in other threads on this topic

[pmiguel]

I posted in another thread a possible work-around:

http://seqanswers.com/forums/showthread.php?t=12523

Briefly, denature (95 oC 2minutes, then "snap cool" on ice until you load it) an aliquot of your sample and run it on a pico chip instead of running a DNA chip.

--
Phillip