Indexing QC distribution

[evan_seq24]

Hi all,

First thread here! My question involves optimizing indexing QC using the Nextera XT library prep kit for DNAseq via Miseq. I'm wondering if anyone can suggest a method for tightening up my indexing QC distribution. I have seen up to 30% CV for samples normalized and run using equal load factors. My goal is to optimize indexing QC distribution to my calculated loading fraction to ensure sufficient coverage on my libraries.

For reference, I perform a picogreen quant for input material, prepare the libraries, and validate fragment size and sample concentration using LabChip. I then normalize to 4nM using those values and proceed with DAL prep. My first thought was that my Caliper reagents or chip were old and generating inaccurate quantification, but I am still experiencing the same issues with a new chip and reagents.

Any ideas are appreciated!

[huguesparri]

In my opinion, 30% CV is not that bad.
The problem is that clustering efficiency is not only depending on library concentration. It also depends on library size as there is a clustering bias toward smaller fragments. So even if you quantify your libraries accurately and mix them in equimolar (or so) proportion, you will face some variations in your final sequence equilibrium as your library profiles are differents.
You could add a size selection step but even if it improves your sequence number equilibirum, it may also reduce you library diversity. This last point can be an issue if you are working on complex/large genomes.

[nucacidhunter]

Two points:
1- your quantification method accuracy is low and variability high so you need to use qPCR if you want more accurate quantification
2- library fragment size distribution as was pointed

[evan_seq24]

Thank you both for the thoughtful responses. Looking at my data, I do see a direct correlation between fragment size and clustering efficiency. I appreciate it!