Because you don't expect to obtain intact total RNA from this sample type, the RIN score (which looks at the 18S:28S ratio to determine RNA integrity) is not meaningful for you. So don't worry about your RIN being low :)

The peak at 25nt is not miRNA. It is the marker for the BioAnalyzer which it uses to mark the starting point of the trace. The RNA Pico kit doesn't have the greatest resolution in this very small size range (your miRNAs would be obscured by your marker peak, for instance) so if you want to assess your small RNA fraction in more detail, it would be worth getting a small RNA BioAnalyzer kit. If you do this, read the protocol carefully because it's weird (eg uses the same priming station settings for DNA chips rather than other RNA chips.)

It's a little difficult to tell from the trace what the other material ~25-200nt is. I'm attaching an image (sorry for low resolution) showing nuclear and cytoplasmic BioAnalyzer small RNA traces from my sample type (mouse brain) - I think the big peak at ~60 is tRNA. My samples are extremely repeatable - traces from different samples look virtually identical. If your results are consistent, I would lean towards assuming that your trace represents the type and size distribution of small RNAs that are present in serum samples. If samples look quite different from each other, it might be that this small material represents greater or smaller quantities of fragments originating from larger, degraded RNAs.

I've used ligation-based small RNA library prep kits extensively, I really like the NEBNext kit. It is very effective at enriching for interesting small RNAs - it captures and ligates miRNAs and snoRNAs with high efficiency, and piRNAs too if the appropriate protocol modification is followed. It will not capture intact tRNAs. Fragmented larger RNAs present in the sample will not be captured if they are chemically fragmented, but will be captured if they were produced by nuclease activity. You can gel-cut your library to sequence just the miRNAs.

If you're thinking about qPCR I suggest using other miRNAs or a panel of miRNAs as a reference gene. Many papers use small nuclear RNAs, snoRNAs etc. as reference (housekeeping) genes to normalise their qPCR. You would probably not expect these to be present at consistent levels in serum samples.

Hope this helps! :)

Laura Leighton , thanks for your response. Part of the issue here is I am sending the samples off for sequencing, it is not something I do in house. So in terms of QC prior to shipping the samples off, that's where I am stuck. Could I do a gel electrophoresis and if there is a band at 25 nt I'm set? Because this is serum, I'm not expecting large RNA, but other small RNA are certainly possible, as you stated. So if there are bands elsewhere on the gel (> 25 nt), does that mean there are other small RNA in the sample or that something is contaminated or degraded and could be creating false miRNA? I am only interested in the miRNA, so if there are peaks elsewhere, my only concern would really be degradation products that look like miRNA at 22 nt but are not really true miRNA.

Janet Grimes I don't think you would see a band at 20-25nt even if you have plenty of miRNA for sequencing - as you can see from the traces in my last post, I can't really see miRNAs on my BioAnalyzer traces, and the BioAnalyzer is a lot more sensitive than a gel would be. The libraries made from these samples have very clear representation of miRNAs and sequenced really well. The presence of other bona fide small RNAs in your samples (tRNAs, etc) is not a problem because some of them won't contribute to the libraries (eg, tRNAs are blocked at the 3' end) and the others can be size-selected out - if your sequencing provider knows you are only interested in miRNAs, they should be able to provide this service.

It's hard to know whether your samples contain only genuine small RNAs, or whether they also contain fragments of larger RNAs. I think the most useful piece of information is how consistent they are. If it's all real small RNA, then you would expect to see virtually the same pattern on BioAnalyzer or gel between different samples of the same type. If a significant amount of it is degraded RNA, you'd expect to see the amount and size of degraded material vary considerably between samples. This is the only thing that would be concerning IMHO - the presence of non-miRNA small RNAs should be a non-issue for library generation and sequencing :)