Hi,
I am preparing small RNA libraries for NGS, using the QIAseq kit of Qiagen
Usually I get Canadian concentrations of around 1-10 ng/ul, but in the last batch all of the libraries were
Hi Iddo
preparing samples is THE key point for further having some good results, for microarrays and NGS. Particularly for RNAseq protocols, you should have done a run with DNase to eliminate contaminant DNA. making some additional PCR cycles could be an interesting idea, the problem is that PCR causes biases and you'll have problems comparing all runs.
fred
Yes, but as I said I have unique Molecular identifiers (UMIs) that are supposed to solve the issue of PCR bias, because you eventually count the number of different UMIs (which is not affected by the number of PCR cycles)
Iddo
Sounds interesting. What exactly it is made those UMIs? How does it work?
Pedro
Hi Iddo,
That’s an interesting question – I’d be interested in you trying it out and sharing the results.
As you’ll be using UMIs, PCR bias shouldn’t be much of an issue; that’s a nice place to start. How many cycles do you usually run on your libraries?
Changing this number by adding three cycles might indeed cause some additional bias, particularly if these samples will be directly compared to samples from previous runs.
I would argue, however, that not adding more cycles might cause even more of a batch effect: without additional amplification, both total library sizes and small RNA reads will be significantly smaller/fewer, which also hampers comparability across runs. This can be mitigated by normalization strategies but is likely to still cause issues.
In our small RNA data, we’ve observed the number of target reads to play a huge role in introducing batch effects between individual sequencing runs…
Best,
Dominik
- Badges
- Science topic
- Similar topics
- Chemistry
- Biochemistry
- Biomolecules
- Nucleic Acids
- RNA
- Small RNA