Highly sensitive and accurate methods are needed to define target allele frequencies (AFs) in next-generation sequencing (NGS) reference material which is critical for the development and characterization of accurate and precise assays for a variety of applications.
Droplet Digital PCR (ddPCR) offers sensitive quantification of AFs, but there is little evidence confirming the concordance between data generated by NGS and ddPCR.
This application note demonstrates the concordance of these technologies for variant calling.
Download this application note to explore:
- Variant detection for SNVs, INDELs and CNVs
- ddPCR as a low cost and easy-to-use solution for AF quantification
- Orthogonal validation of AFs between methods
Droplet Digital PCR
Concordance between Droplet Digital"PCR and Next-Generation Sequencing
in Measuring Allele Frequency for Genomic Reference Materials
Introduction
Genomic reference material (RM) for next-generation sequencing
(NGS) has been widely utilized in clinical diagnosis, NGS product
development, and patient selection for clinical trial or treatments
(Sussman and Rosenbaum 2020). To better serve these purposes,
RM is expected to have precise allele frequencies (AFs) for all
variants contained within. Several methods have been used to
quantify AFs during RM development, including Droplet Digital
PCR (ddPCR") and NGS. At LGC Clinical Diagnostics, ddPCR has
been chosen to quantify and finalize the AFs in our genomic RMs
because of its proven sensitivity, ease of use, accuracy, and low
cost (Whale et al., 2017). Customers report consistent satisfaction
with the RMs and associated AFs claimed, but as yet data directly
comparing the two technologies are limited. Here, we describe a
study that analyzed data from two different types of genomic RMs
and demonstrate that NGS shows a high level of concordance to
variant calls generated with ddPCR as the AF-defining method.
Materials and Methods
Data were generated from the Seraseq Myeloid Mutation DNA Mix
(SeraCare Life Sciences Inc., catalog #0710-0408) using both
ddPCR and NGS technologies and compared at the level of
individual AFs. During product manufacturing, ddPCR quantitation
of blended variant AFs were performed using the QX200" Droplet
Digital PCR System (Bio-Rad" Laboratories, Inc., #1864001).
NGS analysis was performed using the VariantPlex Core Myeloid
Panel (Invitae Corporation) on a MiSeq® System (Illumina®, Inc.,
#SY-410-1003). The Seraseq Myeloid Mutation DNA Mix contains
23 clinically relevant variants across 16 genes, including both single
nucleotide variants (SNVs) and insertion/deletion variants (INDELs)
(Suppl. Table 1). Two CEBPA and one MYD88 variants contained
within the panel were not assayed by either ddPCR (CEBPA, high
GC content) or NGS (MYD88, not covered in the VariantPlex Panel)
and were excluded from the analysis. To test differences in low-level
variant detection between NGS and ddPCR, we analyzed variants
from the Seraseq ctDNA Complete Mutation Mix AF5%, AF2.5%,
AF1%, AF0.5%, and AFO.1% (SeraCare, #
Care, #0710-0528, 0710-0529,
Results
Side-by-side comparison of AFs detected by NGS and ddPCR
at 5 or 10% illustrate that NGS is capable of detecting variants
at these levels and those close to target AFs defined by ddPCR
(Figure 1A, B). The correlation of AFs detected by these two
methods was high (R2 = 0.73). At AF levels of 0.1-5%, all variants
were detected by NGS, which demonstrated high concordance
with ddPCR again (Figure 1C). Together, these results confirm that
AFs measured using NGS have high concordance to AFs defined
by ddPCR, ranging from 0.1-10% in RMs.
Previous studies have noted that NGS tends to underestimate
ddPCR defined AFs (Figure 1B) for complex variants such as FLT3
internal tandem duplication (FLT3-ITD) (Kennedy and Smith 2020)
and NPM1 p.W288Cfs*12 (Akabari et al., 2022). The myeloid data
suggest that both methods are capable of detecting SNVs close
to target AFs (Figure 2A). For INDELs, while the detected mean
AFs using NGS was close to the target AFs defined by ddPCR,
there were several outliers identified, such as FLT3-ITD and NPM1
p.W288Cfs*12 (Figure 2A, B). Similarly, we showed that at even lower
AFs (0.1-5%) NGS detected various types of variants in the Seraseq
ctDNA Complete Mutation Mix, including SNVs, INDELs, and copy
number variants (CNVs), at close to target AFs (Figure 2C). Overall,
these data indicate that NGS can detect variants at target AFs, and
are concordant with AFs defined by ddPCR, across the range of all
variant types examined.
Finally, we examined batch-to-batch variation of AFs for the
Seraseq Myeloid Mutation DNA Mix measured using ddPCR
and NGS. At the AF levels of 5 or 10% we found no significant
difference among the four batches measured using either ddPCR
or NGS (Figure 3).
Target AF, %
Fig. 3. Allele frequencies of variants in different batches of the Seraseq
Myeloid Mutation DNA Mix measured using ddPCR or NGS. ddPCR (.):
NGS (+). AF, allele frequency; ddPCR, Droplet Digital PCR; NGS, next-generation
sequencing.
Conclusions
In summary, our results demonstrated that measurement of
allele frequencies using ddPCR and NGS in DNA RMs show high
concordance from AF 0.1 to 10%. And both methods detect variants
at equivalent levels for all variant types examined, including SNVs,
INDELs, and CNVs. Based on these observations it is reasonable
to conclude that AFs measured using one method (ddPCR) can
orthogonally validate AFs measured using another method (NGS).
It is also possible that genomic RMs with AFs measured using
ddPCR can be used to assess the performance of NGS methods
in the absence of biological samples. Therefore, well-characterized
genomic RMs can provide critical and invaluable information for
optimizing an NGS assay or NGS informatics pipeline.
References
Akabari R et al. (2022). Technological advances: CEBPA and FLT3 internal tandem
duplication mutations can be reliably detected by next generation sequencing.
Genes 13, 630.
Kennedy VE and Smith CC (2020). FLT3 mutations in acute myeloid leukemia:
Key concepts and emerging controversies. Front Oncol 10, 612880.
Sussman R and Rosenbaum JN (2020). Development and validation of molecular
assays for limited tissue samples. Acta Cytol 64, 147-154.
Whale AS et al. (2017). International interlaboratory digital PCR study demonstrating
high reproducibility for the measurement of a rare sequence variant.
Anal Chem 89, 1724-1733.