problem with pooled library cluster generation MiSeq

[subxero]

When pooling 8 TruSeq RNA libraries together for a 50 cycle QC run on the MiSeq I am having a problem in that my libraries do not cluster equally.

I have 2 libraries that both account for ~40% of the clusters each for a total of 80% of clusters and the remaining 6 libraries each account for less than 5% of clusters.

What are possible causes of this?

I QC with BA & Kapa kit and everything looks reasonable from there.

All my other scores for the 50 cycle run look decent enough, 800k clusters/mm^2, 95% PF, 96% >Q30, 86% align.

[GenoMax]

Are the insert sizes similar for these libraries? Perhaps smaller insert libraries are preferentially forming clusters?

[subxero]

the libraries were all prepped the same day with same protocol in same lot, so they should all be quite similar in size

[kmcarr]

never mind