Excessive adapter dimer in Chip input samples but not in IP samples

[mia2025]

Hello all. I have been coming across an issue during my chip seq library prep that has persisted for months now. I am preparing ChiP seq libraries using the Kapa Hyper Prep kit and TruSeq adapters (5uL/reaction) using 20ng DNA from either my Input or IP of transcription factors. The bioanalyzer traces of the libraries look great, however, whenever I quantify using the Kapa QPCR Library quantification kit, my Input samples always show a "double peak" which I think is adapter dimers based on their proximity to the "adapter dimer peak" in the standards melt curve. Strangely enough, I do not have the same issue with my transcription factor libraries. See the attached bioanalyzer trace and Kapa melt curve.

During sequencing, I consistently get good results from my transcription factor libraries, but it is always a mess for my input samples. Does anyone know what this other "peak" is in the kapa melt curve, how to potentially fix it, or how to avoid this in the future?

I am having a hard time really believing it is adapter dimers since i am using the same concentration to generate my TF libraries, and they always look fine.

Any help would be great

Thanks!

[nucacidhunter]

I do not see any adapter-dimer in library trace. qPCR melt analysis is not required for library quantification. Two melt peaks in input library could be due to high diversity of the library while the Chip libraries are less diverse as expected.

I wonder if you could you elaborate what “mess”” means in this context.

[mia2025]

Hi nucacidhunter, thanks for your reply. I agree with your assessment about the lack of dimers in the input bioA trace. To clarify, whenever I sequence libraries that have a similar melt curve to that seen with the input in the attached image, the quality is always very poor. I have uploaded an example of the QC for both the TF and input that were sequenced. As you can see, the TF sequenced (and subsequently aligned) very well, while the input quality was incredibly poor (average QC score below 27).

So, I am still at a loss. Any help or insight would be greatly appreciated.

Thanks again

[nucacidhunter]

It will be helpful if you could post whole FastQC report for input library. Also the sequencing platform, read number and if it was sequenced by itself or multiplied with other libraries.

[mia2025]

Sure, ive attached the QC report for the Input. It was sequenced on the HiSeq SR50 with single read clustering and 1 x 50 sequencing cycles. It was multiplexed with 5 other libraries (including the TFs I referenced above which sequenced very well). In the end, I wound up with only about 9.1M reads.

thanks

[nucacidhunter]

FastQC results indicates relatively low diversity library and containing lots of duplicates which is unusual. I suspect DNA sample prep or library prep is the cause. If you are using the same quantity of DNA as the Chip for library prep, something in the sample is preventing conversion to adapted library fragments.

[Markiyan]

Any residual DNA-ase in CHIP sample?

Is there any residual DNA-ase left in the CHIP seq DNA sample, after the DNA digestion?

If yes, it can chew back the adapter ends, allowing some of them to ligate into the dimers...

[mia2025]

Hi all, thank you all for the replies. I actually figured out the issue. There was a small amount of high MW dna (above 5kb) in all of my inputs that was not visible on an agarose gel but I found after running it on the BioA. I used SPRI beads to remove it all and reprepped the library. It removed that double peak i found in the kapa melt curve and i am getting ready to sequence. I attached a picture of the melt if anyone is curious.

thank you all for the suggestions and comments.