Robust Quantification of Next Generation Sequencing Libraries using Absolute Q Digital PCR

Highlights

  • Consistent reagent partitioning in to greater than 99% of expected partitions compared to 60.9% on emulsion dPCR platform
  • Improved separation of positive and negative partitions in 2-dimensional threshold view
Download PDF Technical Note

Introduction

Advances in next-generation sequencing (NGS) technologies had accelerated the discovery of actionable genomic targets. Accurate quantification of the final library products is critical to maximizing both data quality and output. However, conventional quantitative PCR (qPCR) methods commonly used to assess NGS library concentrations do not evaluate the concentration of complete library fragments. Moreover, in some cases, the amount of final library product is limited, and the input requirement for consistent qPCR quantification can become a hindrance. 

Multiple studies have demonstrated that utilizing digital PCR (dPCR) as a quantification tool for NGS libraries before sequencing helps optimize sequencing run performance, data generation, and data quality1,2. Digital PCR, which uses absolute rather than relative quantification, has superior accuracy and precision at low concentrations relative to qPCR. By leveraging multiplexed digital PCR reactions, users can identify and quantify specific library fragments representing sequenceable molecules (Fig 1).

In this technical note, the Absolute Q Digital PCR Platform was used to perform NGS library quantification of sequenceable library molecules. Four NGS libraries with various insert sequence lengths were quantified in this study.

diagram of NGS library quantification by dPCR

Materials and Methods

Library preparation, separation and dilution

A single mixed-size ATAC-seq library was separated based on fragment length (BluePippin) to create four separate NGS libraries with average fragment lengths of 300, 500, 700, and 1000 base pairs. After size separation, each library was amplified using Q5 polymerase (NEB) and cleaned up using SPRIselect (Beckman Coulter). The amplified libraries were then diluted at 1:200,000 to create the “Dilution 1” then subsequently diluted serially 1:4 for a total of 6 dilutions for each of the 4 fragments lengths

Library preparation, separation and dilution

Digital PCR quantification of the NGS libraries was performed using 5µL of the diluted NGS library per reaction. A detailed breakdown of the digital PCR reagents is listed in Table 1. After preparing the dPCR mix, 9µL of the reaction mixture was loaded into a well of the MAP16 plate, followed by an overlay of 15µL of isolation buffer. The prepared MAP16 plate was then loaded on the Absolute Q. The following thermal parameters were used for each digital PCR run: 96°C hold for 10 minutes, 45 cycles of 95°C denaturation for 5 seconds, 61°C annealing and extension for 30 seconds. Data were collected using the FAM and HEX optical channels.

NGS library quantification reagent preparation

Analysis on Absolute Q

Combinati Absolute Q Analysis Software was used to calculate the concentration of the NGS libraries for each dilution series. The software reports the concentrations of targets that are FAM positive only, HEX-positive only, and FAM/HEX double-positive. By design, partitions containing complete and sequenceable library fragments are positive for both the P5 (FAM) and P7 (HEX) probes. The double positives are represented as a cluster in the upper right quadrant (green dots) of the two-dimensional fluorescence scatter view (Figure 1b). The concentration of double positives was then used in the following equations to determine the concentrations of the original library stock for each dilution point:

Results

Quantification of sequenceable library fragments

The four libraries, which vary in fragment length, were each quantified on the Absolute Q across six serial 4-fold points. The reported concentration in copies/µL from the Absolute Q for the dilution series for each library with specific fragment size are shown in Figure 2. As expected, the observed concentration decreased by approximately 4-fold between each point of the dilution series across all conditions. Using the reported concentration and Equations A and B, the original concentration of each NGS library were calculated to be 2.65nM (±0.78), 3.06 nM (±0.74), 3.06nM (±0.74), 4.63nM (±1.22) for the 300, 500, 700, and 1000 bp NGS libraries respectively (Table 2).

Advantages of microfluidic array partitioning for consistent reagent partitioning

An additional benefit of using Microfluidic Array Partitioning (MAP) technology to perform dPCR is the robustness of reagent partitioning. MAP technology leverages a micro-molded plastic dPCR plate with fixed volume arrays and ensures partitioning across >95% of the available partitions is achieved for all dPCR reactions.

To compare the partitioning efficiency between MAP and emulsion-dPCR, the four NGS libraries were quantified using an emulsion-based dPCR platform in parallel with the Absolute Q. To compare the performance and consistency of reagent partitioning between the Absolute Q and an emulsion-based dPCR platform, the accepted partition or droplet count was compared across 46 paired dPCR reactions. These NGS library quantification reactions were run in parallel on each platform using the same assay and library template material.

Out of a total of 20,480 fixed partitions per MAP dPCR reaction array, the mean accepted partition count for the Absolute Q reactions was 20,412 (±127 partitions). Out of the 20,000 droplets expected per reaction, the mean accepted droplet count for the emulsion-based dPCR reactions was 12,138 (±1267 droplets). For this dataset, the minimum number of accepted partitions for Absolute Q reactions was 19,645 and the minimum number of accepted droplets for the emulsion-based dPCR reactions was 8,629 (Figure 3).

Finally, a comparison of the representative 2-dimensional partition fluorescence plots between the emulsion-based dPCR platform (Figure 3b) and Absolute Q dPCR (Figure 3c) highlights the improved signal separation between the resulting double-positive partitions and negative partitions. The improved separation aids in consistent thresholding of dPCR data and as a result more robust and reproducible quantification results.

Summary

Quantification of NGS libraries by digital PCR is advantageous because it is possible to distinguish complete sequenceable library molecules and perform absolute quantification. Given that the final concentration of NGS libraries can vary based on the preparation method and performance of the method used, it is critical for each digital PCR reaction to be as consistent as possible to maintain robust quantification across a wide range of input concentrations.
The highest level of precision for dPCR quantification occurs when an average of 1.59 copies of target per partition is present in the reaction volume3. This means for high concentration samples the total number of accepted partitions is critical because more analyzed partitions can improve precision by bringing the average closer to 1.59. In the case of NGS library quantification, because the sample consists of amplified nucleic acids, the target concentrations are usually very high. The consistently high analyzed partition numbers of the Absolute Q ensure high levels of precision even at high target concentrations, ensuring high-quality NGS library quantification.
Here we demonstrated the capabilities of the Absolute Q to perform robust NGS library quantification across a wide range of fragment sizes and concentrations while achieving an exceptionally high number of accepted partitions per dPCR reaction.

References

  1. Robin, Jérôme D., et al. “Comparison of DNA Quantification Methods for Next Generation Sequencing.” Scientific Reports, vol. 6, no. 1, 2016, doi:10.1038/srep24067.
  2. White, Richard A, et al. “Digital PCR Provides Sensitive and Absolute Calibration for High Throughput Sequencing.” BMC Genomics, vol. 10, no. 1, 2009, doi:10.1186/1471-2164-10-116.
  3. Majumdar, Nivedita, et al. “Digital PCR Modeling for Maximal Sensitivity, Dynamic Range and Measurement Precision.” PLOS ONE, vol. 10, no. 3, 2015, doi:10.1371/journal.pone.0118833.

More to Explore