Question: How do alternative quantification methods other than qPCR impact library loading?
Answer: Library quantification is a critical step for achieving high-quality sequencing. The loading concentration recommendations for 10x libraries listed in the Sequencing Section of the User Guides and Sequencing Metrics & Base Composition Technical Notes were developed using KAPA Library Quantification Kits to determine optimal library concentration prior to Illumina sequencing. While under clustering can maintain high data quality, it results in lower data output. Alternatively, overclustering can lead to run failure, poor run performance, lower Q30 scores, introduction of sequencing artifacts, and lower total data output.
Alternative quantification methods such as Qubit or Bioanalyzer may report different concentrations than KAPA qPCR due to inherent differences in the technologies. Due to these differences, the use of alternative quantification methods may lead to sequencing runs that are over or under loaded. Different library types due to their different fragment sizes may be more or less affected by these differences in quantification methods. Additionally, different sequencing platforms may be more or less sensitive to loading concentrations.
If you plan on using an alternative quantification method, we strongly recommend performing a comparison between KAPA qPCR and the method of choice before performing an initial sequencing run with the specific library type that will be sequenced (see Figure 1 as an example). It is important to perform this comparison using your own instruments and reagents as the results may vary. This comparison can then be used to adjust the amount of library loaded.
For example if a library quantified by KAPA qPCR measures as 300 nM and the same library measures as 200 nM by Qubit, then loading the library at the recommended concentration for the sequencer based on the Qubit quantification would result in loading 50% more library than if the KAPA qPCR value had been used. Overloaded sequencing runs may lead to overclustering on the flow cell. Overloading/overclustering can lead to poor run performance, lower Q30 scores, possible introduction of sequencing artifacts, and, counterintuitively, lower total data output (i.e., fewer usable reads) as compared to optimally loaded runs (Figure 2). Overloading does not result from instrument issues or sequencing consumable issues, but rather from loading too much library.
As a general note, the loading recommendations for an individual sequencer are listed in the corresponding Sequencing Section of the User Guides and Sequencing Metrics & Base Composition Technical Notes as an initial starting concentration. Additional optimization for optimal loading concentrations may be required.
Please see the following Knowledge articles from Illumina for additional guidance:
- Cluster density guidelines for Illumina sequencing platforms using non patterned flow cells
- Plotting %Occupied by %PF to optimize loading for the NovaSeq X/X Plus, NovaSeq 6000, and iSeq 100
- Calculating Percent Passing Filter for Patterned and Nonpatterned Flow Cells
- Diagnosing Suboptimal Clustering (Patterned Flow Cells)
- How short inserts affect sequencing performance
- Best practices for library quantification
- Sequencing Chemistry
Figure 1) Comparison of final library concentration as measured by KAPA qPCR and Qubit. Gene Expression libraries for various 10x assays were quantified using Qubit™ 1X dsDNA High Sensitivity Assay (ThermoFisher, Q33231) or by qPCR using the KAPA Library Quantification Kit for Illumina Platforms (Roche, KK4824). The average size of the libraries was calculated using LabChip using the DNA High Sensitivity Regent Kit (Perkin Elmer, CLS760672) over 150-300 bp or 200-2000 bp windows for the Flex and CytAssist Gene Expression libraries, or Single Cell 3’ Gene Expression libraries, respectively. Qubit systematically results in lower reported concentrations than KAPA qPCR, and can result in overloading Flowcells (see Figure 2).
Figure 2) - Single plexity Flex libraries loaded on a NovaSeq6000 SP Flowcell on separate lanes using the Xp workflow. Lane 1 was loaded with standard conditions (10% PhiX at 150 pM), and Lane 2 was deliberately overloaded with low PhiX (1% PhiX at 300 pM). A) %Occupied by %PF plots were generated confirming the overloaded phenotype in Lane 2 (red) and optimal/underloaded phenotype in Lane 1 (blue). B) Demultiplexing report highlights >93% of reads on Lane 1 were able to be demultiplexed, while only 20% of Lane 2 were able to be demultiplexed. %Base plots for Lane 1 (C) and Lane 2 (D) are shown. The overloaded sample on Lane 2 results in an abundance of polyG in i5, leading to the lower demultiplexing efficiency. The single plexity sample loaded at the recommended concentration does not result in polyG in i5 and sample is able to be demultiplexed efficiently.
Products: Single Cell Gene Expression, Single Cell Gene Expression Flex, Fixed RNA Profiling, Single Cell Immune Profiling, Single Cell ATAC, Single Cell Multiome ATAC+GEX, Visium for fresh-frozen, Visium for FFPE, Visium CytAssist Gene and Protein