Spiky library distribution on HS DNA chip

[sk8bro]

Hi,

I have made strand-specific RNA seq libraries. Some basic info is 500 ng total RNA (estimated from nanodrop) was inputted to a Turbo Dnase reaction, followed by Ribozero, fragmentation, first strand synthesis, dUTP strand-specific second strand synthesis and library prep, and 13 PCR cycles.

I realize the T16 sample is overamplified and that I need to do another clean-up to remove the adapter dimers, but what I wonder is if anybody has any insight specifically on is the jagged, or 'spiky', look of my final amplified libraries as assessed by HS DNA chip? I haven't seen this in my past libraries, although these past good libraris were made from 5 ug of different RNA and I modified my protocol a bit with this last batch.

I've also included intermediate BioA traces of the Ribozero'd RNA, fragmented RNA, and First Strand Synthesized RNA:cDNA hybrid.

My best guess is too low of input after the Dnase and various clean ups....

Thanks, K

[pmiguel]

My concern would be that the big spikes are concatamers of the ~120 bp adapter dimer. Shouldn't happen, but maybe if there is a single-stranded nuclease degrading the Y-adapter ends, you could get that.

We have sequenced spikey samples like this when there just wasn't any more RNA sample available and the user really wanted data. Results are generally not great, sometimes only 10% of the reads come from the sample. The other 90% from, well who knows? A variety of hits to human and bacteria, in general. I have always interpreted this as deriving from trace contamination of library construction reagents (enzymes?) that normally would be swamped out by the sample RNA. But, if, as you suggest, your input is too low, the minor contaminants could predominate via the magic of PCR.

--
Phillip