Reduce bias in small RNA library prep with randomized adapters

[Bioo Scientific]

The NEXTflex™ Small RNA Sequencing Kit v2 provides an easy, flexible, cost-effective solution for generating libraries with reduced bias from small RNA using Illumina sequencing platforms. Bioo Scientific’s proprietary approach to overcoming ligation bias in small RNA-seq libraries involves using a pool of adapters having randomized sequences at the ligation site, where each adapter sequence is present in vast molar excess over any given small RNA in the sample. Most of the bias in adapter ligation is due to the sequence of the nucleotides adjacent to the target junction. No single adapter sequence is able to efficiently ligate to all small RNAs, indicating that the target sequence, as well as the adapter sequence, is a source of bias. The new NEXTflex Small RNA Sequencing Kit v2 uses a randomized adapter strategy to allow small RNAs of any sequence to “find” their corresponding optimal adapters, resulting in small RNA-seq libraries that show a dramatic reduction in bias.

[Bioo Scientific]

Download Bioo Scientific’s whitepaper describing how randomized adapter reduce ligation bias in Illumina-compatible small RNA-seq libraries.

[nucacidhunter]

I wonder if Bioo scientists have any comment on this publication:

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0126049

[Bioo Scientific]

Hi Nucacidhunter,

Thanks for the question. The work by NEB is certainly interesting. It looks like the only original findings are that bias can also be reduced by using either terminal or internal random bases and that efficiency can be improved through co-folding adapters. Other conclusions, like both 3’ and 5’ ligation contributing to bias and the observation that some bias can be explained through secondary structure and co-folding, are mostly just confirmations and more detailed analysis of previous findings. It should also be noted that all of these experiments were performed on a synthetic miRNA pool, so it will be interesting to see how this strategy works with total RNA samples. The strategy implemented for avoiding adapter-dimer formation will be particularly interesting, as adapter-dimer is a much bigger problem with libraries prepared from real-world samples and is more difficult to prevent when using adapters with randomized regions. It will also be interesting to see if other, independent groups confirm these findings, as is the case with the bias reducing effects of using adapters with randomized ends.

One thing that caught our eye was the claim that “Randomized adaptor regions adjacent to the ligation junctions can further complicate sequence analysis if an unknown sample were to be sequenced because the termini of the sRNA in a sequencing read cannot be precisely determined.” As long as the read goes into the adapter, the random bases are easily identified and removed as the first 4 bases of the read and the last 4 bases of the read prior to the adapter sequence (assuming 4 nt random regions were used), allowing precise determination of sRNA ends. And if the insert is too long for the read to go into the adapter sequence, the 3' end cannot be determined no matter whether the random bases were at the ligation junction or internal.

In fact, internal random bases are more likely to complicate sequencing than random bases at the ligation junctions, as the constant region that is read at the beginning of the reads can cause issues related to low diversity. This was illustrated in NEB’s publication when the authors noted that “Two sets of new adaptors (v1 + v2) were used for Illumina sequencing in order to generate sequence diversity for the instrument in the 7 base region between the randomized region and RNA insert...” The only other comparison of randomized ends versus internal randomization in reducing bias concludes that “Surprisingly, the improvement in sequencing results was similar when either the 3’ Rand or 3 ’MidRand adaptor were used.” Thus, it appears that NEB’s internal randomization strategy does not provide any advantages over end randomization.

At Bioo Scientific we have long recognized the importance of reducing bias in small RNA library prep, resulting in last year's launch of the NEXTflex™ Small RNA Sequencing Kit v2, which uses randomized adapters to reduce ligation bias. We followed that up with the recent launch of the NEXTflex™ Small RNA Seq Kit v3, which combines bias reduction with completely gel-free or low-input protocols. These kits are currently available, providing scientists with optimized kits for reduced-bias small RNA library preparation.

[nucacidhunter]

Thanks for your expert comments and update on kit.

[mama]

We are currently using the Nextflex small RNA Seq kit v2 and we get huge amount of adapters dimers in our libraries. We use a PAGE TBE 10% gel after the PCR and most of the time the band at 150 bp is not as sharp and tight as Bioo scientists says. We retrieve this band at 150 bp and up to 80% of this band is adapters dimers (130-140bp). It means that adapters dimers could also migrate at 150 bp.

[Bioo Scientific]

Hi Mama,

It is unusual to see such a high percentage of adapter-dimer product in your libraries. We would be happy to help you troubleshoot this. Can you send the following information to our tech service group at [email protected]:

What was the input material type and amount?
How was the input material isolated?
How many cycles of PCR were performed?
Do you have gel and/or Bioanalyzer images of post-PCR samples and final libraries?
Was a library prepared in parallel with the included miRNA control?