Library Preparation > General > Troubleshooting

• Adapter dimers causes, effects, and how to remove them

• Library QC and Troubleshooting with the Bioanalyzer and Fragment Analyzer Support Webinar Video

• Preventing Contamination Expert Video Tip Video

• Troubleshooting elevated PhiX alignment in sequencing runs

**What are adapter dimers?**During quality checks of sequencing libraries with chip-based capillary electrophoresis instruments, eg, the BioAnalyzer or Fragment Analyzer, or on agarose gel, an unexpected small peak at 120-170 bp indicates the presence of adapter dimers (Figure 1).

Adapter dimers contain full-length adapter sequences that are able to bind and cluster on the flow cell and generate sequencing data. In contrast, primer dimers do not contain complete adapter sequences, and are not able to bind or cluster on the flow cell, so are not sequenced.

What causes adapter dimers?

Insufficient starting material

• Using too little input material can lead to an increase in adapter dimer formation. Quantification with a fluorometric-based method is recommended to ensure accuracy of input amount. Using an input amount within the range recommended for the workflow minimizes the chance that adapter dimers will be present in the final libraries.

Poor quality of starting material

• Using fragmented or degraded input nucleic acid is not recommended for certain workflows. Using degraded input material for library preparation methods that are not compatible with degraded input can lead to adapter dimers.

Inefficient bead clean-up

• It is important to follow best practices when handling beads to ensure proper size selection and removal of adapter dimers that may have formed during library preparation.

Why is it important to remove adapter dimers?

Adapter dimers have complete adapter sequences, so they can cluster and sequence. Because of their small size, they cluster more efficiently than the intended library fragments. Depending on the quantitative proportion relative to the specific library, they can subtract a significant portion of sequencing reads from the desired library fragments. Moreover, they can negatively impact sequencing data quality, and may even cause a run to stop prematurely (Figure 2).

How to remove adapter dimers?

If adapter dimers are present in the library, perform an additional clean-up step with beads (AMPure XP, SPRI, or Sample Purification beads, “SPB”) or gel purification. A second round of purification may reduce the library yields. A bead ratio of 0.8x to 1x is usually recommended and sufficient to remove the unwanted adapter dimers.

What do adapter dimers look like?

Figure 1. BioAnalyzer electropherogram showing a library with an adapter dimer peak at 126 bp.

Figure 2. If adapter dimers are present during sequencing, this is evident in the percent base (%base) plot in Sequence Analysis Viewer, or in BaseSpace. Adapter dimers give the following signature:

1. A region of low diversity

2. The index region

3. Another region of low diversity

4. An increase in "A" when the adapter sequence ends and the sequencing read runs into the flow cell. This A overcall may be a G overcall, depending on the sequencing platform used

The overcall pattern may differ depending on the proportion of adapter dimers compared to the intended library fragments, ie, it can be less recognizable with a lower proportion of adapter dimers present in the run.

Note: Patterned flow cells are more susceptible to reduced run performance when adapter dimers are present. Illumina recommends limiting adapter dimers to 0.5% or lower when sequencing on patterned flow cells and 5% or lower when sequencing on non-patterned flow cells. (Percentages refer to the percent of adapter dimers relative to the full library; this can be determined by regional analysis with trace/electropherogram instruments.) Any level of adapter dimers will take away reads from the proper library fragments.

This support webinar video addresses how to use the Agilent Bioanalyzer to check library quality prior to sequencing, and to troubleshoot sample preparation. The video covers the following topics: how to identify features of an ideal library trace, how to recognize potential issues, and how to prevent potential issues.

Video link: https://youtu.be/uie_UYNjFSk

Video length: 40:56 min

Many next-generation sequencing applications use the Polymerase Chain Reaction (PCR) for DNA amplification. This video presents best practices to minimize the potential for PCR contamination in experiments.

<

Video length: 10:12 min

​​​​​​​

The PhiX Control v3 Library (commonly referred to as PhiX, FC-110-3001) is commonly added to sequencing runs as a positive control for sequencing performance or to help balance low-diversity libraries. The addition of PhiX to a sequencing run also allows for additional sequencing metrics to be calculated during a run, enabling deeper analysis of loading optimization when comparing the %Align metric to the expected PhiX Spike-in by volume.

Occasionally, users may find that the %Align metric reports a significantly higher value than was spiked-in. Unlike low PhiX alignment which can have multiple root causes, high PhiX alignment is the result of PhiX taking up a larger than expected proportion of the sequenced material on a run. There are several possible causes for this phenomenon.

1. Inaccurate quantification of the library and/or PhiX.

2. Too low of a loading concentration used for the library pool.

3. Issue with library design resulting in poor clustering or priming.

Sequencer or sequencing consumable performance issues impact libraries and PhiX alike, so elevated PhiX alignment is not indicative of instrument malperformance or poor reagent quality.

Interpreting sequencing results to determine causes of high PhiX alignment

Users can determine which of the possible causes are resulting in high PhiX by examining the sequencing metrics of their run. If a run has high Q30 scores, low error rates, and either low density (if sequencing with a non-patterned flow cell like MiSeq or NextSeq 500/550 systems) or low %Occupancy (if sequencing with a patterned flow cell like iSeq 100, MiSeq i100, NextSeq 1000/2000, NovaSeq 6000, or NovaSeq X/X Plus systems), this indicates that the run is under-clustered. The root cause in this case is an error in quantification, QC, too low of a library loading concentration, or design which resulted in PhiX taking up a larger proportion of the sequenceable material on the flow cell than expected.

On the other hand, if a run shows lower Q30 scores, high density/occupancy, and higher error rate along with higher than expected PhiX Alignment, this could be an indication of run over-clustering in which PhiX was loaded at a higher concentration than expected and resulted in elevated clustering. Users may also see these phenotypes in combination with low Index 1 quality or a high proportion of Undetermined reads after FASTQ generation, both of which may be a sign of high PhiX presence in a sequencing run.

If a run is indicating >90% PhiX Alignment, this is the result of a total or near-total clustering failure of the library and is related to library design or compatibility issues.

Troubleshooting high PhiX alignment

1. Quantify both PhiX and the library pool to verify accurate loading concentration calculations. Illumina recommendations for quantification methods can be found in the Knowledge article Library quantification and quality control quick reference guide

   1. If the PhiX stock concentration is significantly higher than the expected 10 nM, contact Illumina Technical Support. The PhiX can still be used on future runs, but loading calculations should be adjusted to reflect the measured concentration of PhiX.

   2. If library pool concentrations are significantly lower than expected, adjust dilution concentrations to achieve optimal loading for the sequencer used. Reference the Denature and Dilution Guide for Illumina Library preparation methods.

   3. If quantification is as expected, it may be necessary to use a higher loading concentration for the library pool for optimal clustering.

2. In cases where PhiX Alignment is >90%, libraries should be remade to make sure all required regions of the library adapters are present and that the libraries are able to successfully bind to the flow cell. Make sure that libraries are compatible with Illumina instruments, especially when using custom primers and/or third party kits. More information can be found in the Knowledge article Considerations when migrating non-Illumina libraries between sequencing platforms.

For assistance with any troubleshooting steps, contact Illumina Technical Support.