Improve sensitivity for rare targets by increasing total analyzed sample using digital pooling of digital PCR data


A primary application of digital PCR (dPCR) is rare target detection. Absolute quantification, rather than cycle threshold values enable users to obtain the count of original target molecules in a sample – eliminating the need for interpreting amplification curves or relying on reference material to provide quantitative measurements. This improves both accuracy and reproducibility of concentration measurements of rare targets.
However, the limit of detection for any digital PCR reaction is dependent on several factors. These include the concentration of the target of interest and the volume of sample that can be loaded per reaction (Figure 1). In scenarios where the target of interest is present at a very low concentration, it may be necessary to run more than one reaction in order to sample enough volume to detect the target.

One example of rare target detection is the measurement of circulating tumor DNA (ctDNA) among normal circulating cell free DNA molecules (ccfDNA) from liquid biopsy samples of individuals diagnosed with cancer. In this application, the total number of ctDNA molecules can vary from patient to patient as well as from day to day; and are often very low in concentration – typically only several molecules per microliter. If the concentration of ctDNA molecules is exceptionally low, the amount of sample tested will directly impact the number of ctDNA molecules that can be detected. Furthermore, many existing digital PCR platforms suffer from lack of consistency and reproducibility in the total number of partitions generated due to limitations in the underlying partitioning technology, resulting in unreliable detection. This inconsistency can cause variability in the total amount of sample analyzed per reaction as well as between experiments. The wasted sample can contribute to sub-sampling error. Unlike emulsion based dPCR platforms, the microfluidic array partitioning (MAP) technology used in Absolute Q dPCR partitioning is entirely automated and highly consistent – utilizing an industry leading 95% of loaded sample volume across over 20,000 partitions each time.
This technical note highlights the digital pooling feature of the Combinati Absolute Q by leveraging an Applied Biosystems® TaqMan™ Liquid Biopsy Assay for the cancer mutation PIK3CA p.H1047R (cat. A44177). A contrived sample with mutant allele fraction at 0.1% was created and tested and digitally pooled across 4 dPCR reaction arrays (~81,920 partitions) .

Download PDF Technical Note

Workflow and Methods

DNA Mixture and dPCR reaction preparation

Using genomic DNA (Promega Human Male Control) and plasmid DNA containing the PIK3CA p.H1047R mutation, a DNA mixture of 0.1% MAF was prepared with final human genomic DNA concentration of approximately 1.15 ng/µL. The Applied Biosystems TaqMan Liquid Biopsy assay for PIK3CAp.H1047R which detects both the wild type (VIC) and mutation (FAM) alleles was used in this study. The digital PCR reactions were prepared according to the volumes indicated in Table 1.

Digital PCR reagent table

Reagent Preparation and Digital Pooling

After preparing the dPCR mix, 9µL of the reaction mixture was loaded into the MAP16 plate followed by an overlay of 15µL of isolation buffer. The prepared MAP16 plate was then loaded on the Absolute Q. Standard thermal parameters for use of the Applied Biosystems® TaqMan™ Liquid Biospy Assays on Absolute Q were used. Following the dPCR run, target concentrations were determined using the Absolute Q Analysis Software. A total of eight reactions were run using the samples and subsequently four arrays each were digitally pooled for a total of two replicates.
As an orthogonal test, a positive control at higher concentration of 0.1% MAF DNA mixture was prepared using the same plasmid and human genomic DNA. This DNA sample was used for 2 independent replicates and loaded at a final concentration of 33ng per reaction or approximately 10,000 wild type molecules and 10 mutation molecules per reaction.


After dPCR was complete, all reactions were analyzed using the Absolute Q Analysis software. The dPCR results for the pooled low concentration 0.1% MAF sample across 4 arrays and high positive concentration control are shown in Table 2. Both conditions reported concentrations of the mutation and wild type PIK3CA molecules similar to the expectations. As indicated in Table 2, across approximately 81,545 digitally pooled partitions, the PIK3CA mutation allele was calculated to be at a concentration of 11.5 copies per reaction (cp/reaction) – similar to the positive control reaction which was determined to be 6.9 cp/reaction. Finally, the observed MAF of the digitally pooled low concentration sample was 0.1% MAF as expected.


Digital pooling is an effective method to increase sensitivity by increasing the amount of volume that can be analyzed for a given sample. In this technical note, four arrays (36µL dPCR Mix) of the MAP16 plate were used to analyze a total input sample volume of 28.8µL of a 0.1% MAF DNA mixture. Overall this method has the potential to be applied to a multitude of rare target detection applications in the precision medicine space such as monitoring treatment response, screening for minimal residual disease or rapid identification of mutations linked drug resistance.

More to Explore