{
  "id": "digital-pcr",
  "title": "Digital PCR",
  "lang": "en",
  "source_lang": "en",
  "summary": "Digital PCR quantifies nucleic acids by partitioning a sample into many reactions, classifying each as positive or negative, and estimating target abundance statistically.",
  "category": "science",
  "subcategory": "biology",
  "type": "concept",
  "aliases": [
    "Digital polymerase chain reaction",
    "dPCR"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/digital-pcr/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/digital-pcr/",
      "json": "https://aiwiki.page/zh/digital-pcr.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Digital PCR (dPCR) is a method for quantifying specific nucleic acid sequences by dividing a sample among many separate polymerase chain reaction (PCR) reactions and counting those that produce detectable amplification. Statistical correction for partitions containing more than one target yields an estimate of the original target concentration. Unlike conventional quantitative PCR, digital PCR generally does not require a calibration curve made from standards of known concentration. It is used to measure DNA, quantify RNA after reverse transcription, detect rare sequence variants, and determine differences in gene copy number."
    },
    {
      "heading": "Principle",
      "text": "PCR amplifies a selected nucleic acid sequence through repeated cycles of strand separation, primer binding, and extension by DNA polymerase. In digital PCR, this reaction takes place in thousands to millions of physically separated compartments, depending on the platform.\n\nA sample is mixed with amplification reagents and distributed into droplets, microwells, or other small chambers. Some compartments receive no copies of the target sequence; others receive one or several. After amplification, each accepted partition is classified as:\n\n- Positive, indicating detectable amplification of the target.\n- Negative, indicating no detectable amplification of the target.\n\nDetection usually uses fluorescence from sequence-specific probes or DNA-binding dyes. “Digital” refers to the conversion of these signals into discrete positive and negative outcomes, rather than to the absence of analog measurements: the underlying fluorescence intensity is continuous.\n\nA positive partition does not necessarily represent exactly one original copy. Multiple target copies may enter the same compartment, making statistical correction essential except at very low occupancy."
    },
    {
      "heading": "Statistical basis and quantification",
      "text": "The Poisson model\n\nWhen target molecules are randomly and independently distributed among equal-volume partitions, the number of copies in a partition is commonly modeled by a Poisson distribution.\n\nIf \\(\\lambda\\) is the mean number of target copies per partition, the probability that a partition contains no copies is\n\n\\[P(0)=e^{-\\lambda}.\\]\n\nFor \\(N\\) accepted partitions, of which \\(N_0\\) are negative, the estimated mean occupancy is\n\n\\[\\hat{\\lambda}=-\\ln\\left(\\frac{N_0}{N}\\right).\\]\n\nEquivalently, if \\(p\\) is the observed positive fraction,\n\n\\[\\hat{\\lambda}=-\\ln(1-p).\\]\n\nFor an effective partition volume \\(v\\), the estimated concentration in the partitioned reaction mixture is\n\n\\[\\hat{c}=\\frac{\\hat{\\lambda}}{v}.\\]\n\nThe estimated number of target copies in the accepted partitions is \\(N\\hat{\\lambda}\\). Reporting concentration in the original specimen additionally requires accounting for dilution, extraction, elution volume, and the fraction of the extract tested.\n\nFor example, if 20% of accepted partitions are positive, then\n\n\\[\\hat{\\lambda}=-\\ln(0.8)\\approx 0.223.\\]\n\nThus, 10,000 accepted partitions contain an estimated 2,231 target copies, rather than the 2,000 suggested by simply counting positive partitions.\n\nPrecision and operating range\n\nThe positive and negative counts have sampling variability, which can be used to calculate confidence intervals. The number of accepted partitions and their occupancy strongly influence precision.\n\nAt very low occupancy, uncertainty is dominated by the small number of target molecules sampled. At very high occupancy, few negative partitions remain, making the Poisson correction less precise. If every partition is positive, the basic negative-fraction calculation cannot provide a finite concentration estimate; dilution or a different partitioning arrangement is required.\n\nThe model assumes sufficiently random distribution, known partition volumes, and reliable classification. Unequal volumes, molecular aggregation, physical linkage between targets, or amplification failure can require additional modeling or experimental controls."
    },
    {
      "heading": "Workflow and platform types",
      "text": "Sample preparation\n\nDNA extraction must recover target material while limiting substances that inhibit amplification. For RNA measurements, reverse transcription first converts RNA into complementary DNA. This can occur before partitioning or within partitions in a one-step reaction.\n\nSome assays use restriction digestion or controlled fragmentation to reduce DNA viscosity, improve distribution, or separate linked target copies. Such treatment must avoid disrupting the sequence that the assay amplifies.\n\nPartitioning\n\nTwo widely used formats are:\n\n- Droplet digital PCR: The aqueous reaction mixture is dispersed into droplets within an immiscible carrier fluid. Droplets undergo thermal cycling and are subsequently read individually or imaged.\n- Chamber-based digital PCR: The reaction mixture fills an array of fixed-volume wells or chambers, which are isolated before amplification and usually read by imaging.\n\nPlatforms differ in partition number and volume, sample capacity, automation, optical channels, and the fraction of loaded material ultimately analyzed. Microfluidics is central to many partition-generation systems, although not all implementations use the same fluid-handling architecture.\n\nAmplification and readout\n\nThermal cycling amplifies the target in occupied partitions. Most routine systems classify partitions using endpoint fluorescence rather than the cycle at which fluorescence first becomes detectable.\n\nA common complication is rain: partitions with intermediate fluorescence between the main negative and positive clusters. Rain can result from partial inhibition, sequence mismatches, damaged templates, variable amplification, or technical artifacts. Threshold selection and treatment of intermediate partitions can materially affect results, particularly for rare targets.\n\nMultiplexing\n\nMultiplex assays measure several targets in the same reaction. Different fluorescent labels, signal amplitudes, or combinations of channels can distinguish target classes.\n\nMultiplexing supports target-to-reference ratios, variant detection, and assays of molecular linkage. Its complexity is constrained by optical resolution, overlapping signal clusters, competition among reactions, and the need to validate each target combination."
    },
    {
      "heading": "Comparison with other PCR methods",
      "text": "Conventional endpoint PCR generally establishes whether an amplification product is present, often using gel electrophoresis. Quantitative PCR (qPCR), also called real-time PCR, monitors amplification during cycling and uses the cycle at which fluorescence crosses a threshold to infer starting abundance.\n\nDigital PCR instead estimates abundance from the distribution of endpoint-positive and endpoint-negative partitions.\n\nFeature | Digital PCR | Quantitative PCR\nMain measurement | Fraction of positive partitions | Amplification trajectory and threshold cycle\nTypical quantification basis | Partition occupancy and volume | Calibration curve or relative comparison\nEffect of amplification efficiency | Less dependent on amplification kinetics, provided occupied partitions become detectable | Strongly dependent on amplification efficiency and its consistency\nConcentration range | Limited by partition number, volume, and occupancy | Often broad within a validated assay\nCommon strengths | Copy-number ratios, low-level variants, reference measurements | High-throughput testing and broad-range quantification\n\nDigital PCR is often described as providing absolute quantification, meaning that target concentration can be estimated without an external target calibration curve. This does not mean the result is free of calibration requirements or systematic error. Partition-volume accuracy, dilution, extraction recovery, detection efficiency, and data analysis still affect the result.\n\nNeither method is universally superior. Performance depends on the sample, target, required sensitivity, concentration range, and validated workflow."
    },
    {
      "heading": "Applications",
      "text": "Rare-variant detection\n\nDigital PCR can distinguish a low-abundance mutation from a large background of nonvariant sequence using variant-specific assays. Partitioning reduces the number of competing background molecules within individual reactions.\n\nApplications include research and validated clinical assays involving cancer-associated variants and mosaicism. The lowest measurable variant fraction is not a universal property of digital PCR: it depends on input quantity, assay specificity, background false positives, and the number of informative molecules tested.\n\nCopy-number measurement\n\nA target sequence can be measured relative to a reference sequence to assess gene copy number. Such assays are used to investigate deletions, duplications, and amplifications.\n\nConversion of a target-to-reference ratio into copies per cell requires assumptions about the reference locus, chromosome content, and sample composition. Mixed cell populations or variation in the reference sequence can complicate interpretation.\n\nRNA and gene expression\n\nReverse-transcription digital PCR measures RNA-derived complementary DNA and is used in gene expression studies and RNA-virus quantification. Reverse-transcription efficiency remains a potential source of bias: counting complementary DNA copies does not automatically establish the number of original RNA molecules.\n\nMicrobial and environmental analysis\n\nDigital PCR is used to quantify nucleic acid targets from viruses, bacteria, and other organisms in clinical, food, agricultural, and environmental samples. Applications include wastewater analysis and detection of low-abundance environmental targets.\n\nDetection of a nucleic acid sequence does not by itself establish that an organism is viable or infectious.\n\nReference materials and biotechnology\n\nDigital PCR contributes to assigning values to reference materials, comparing measurement procedures, and supporting metrological traceability. It is also used to quantify engineered DNA constructs, vector sequences, and the outcomes of genome editing experiments.\n\nIt measures predefined target sequences rather than discovering unknown sequence variation. DNA sequencing and digital PCR therefore serve different, often complementary, purposes."
    },
    {
      "heading": "Limitations and quality assurance",
      "text": "Sampling limits\n\nPartitioning cannot detect molecules absent from the aliquot analyzed. For an expected input of \\(\\mu\\) target copies, random sampling alone gives a probability of no target being present of\n\n\\[P(0)=e^{-\\mu}.\\]\n\nIncreasing partition number does not compensate for an insufficient number of molecules in the tested specimen. More partitions can improve occupancy resolution, but low-copy detection also requires adequate sample input.\n\nBackground and detection limits\n\nContamination, nonspecific amplification, probe cross-reactivity, and misclassification can produce false positives. Inhibition, sequence mismatches, damaged templates, and amplification failure can produce false negatives.\n\nThe limit of blank, limit of detection, and limit of quantification describe different aspects of performance: background signal, reliable detection, and quantification with a stated level of precision and accuracy. They must be established for the relevant assay and sample matrix rather than inferred solely from instrument specifications.\n\nMolecular accessibility and linkage\n\nDigital PCR measures amplifiable target units under the conditions of the assay. Fragmented DNA may retain one assay target while losing another. Conversely, linked copies on the same DNA fragment may enter a partition together rather than distribute independently.\n\nThis distinction matters when interpreting results as numbers of genomes, cells, or organisms. A target-copy count is not automatically equivalent to any of those quantities.\n\nMeasurement uncertainty and reporting\n\nA complete assessment of measurement uncertainty extends beyond partition-counting statistics. Relevant contributions include sample heterogeneity, extraction recovery, dilution, partition volume, reverse transcription, and classification rules.\n\nThe Digital MIQE reporting guidelines promote transparent descriptions of sample handling, assay design, partitioning, controls, data processing, and uncertainty. Their purpose is to make results interpretable and reproducible, rather than to certify a particular platform or assay."
    },
    {
      "heading": "Historical development",
      "text": "Digital PCR developed from limiting-dilution PCR, in which samples were diluted across multiple reactions to estimate target abundance from positive and negative results. Work by Sykes and colleagues in 1992 demonstrated PCR-based quantification using limiting dilution. Bert Vogelstein and Kenneth Kinzler introduced the term “digital PCR” in 1999 in work on counting individual templates and detecting rare sequence variants.\n\nEarly approaches required many separately handled reactions. Later advances in microfabrication, droplet generation, fluorescence detection, and automated analysis made large-scale partitioning practical. These developments established digital PCR as a routine quantitative method alongside conventional PCR and real-time PCR."
    }
  ],
  "links": [
    "nucleic-acid",
    "polymerase-chain-reaction",
    "dna",
    "dna-polymerase",
    "fluorescence",
    "poisson-distribution",
    "reverse-transcription",
    "rna",
    "microfluidics",
    "gel-electrophoresis",
    "quantitative-pcr",
    "calibration",
    "mutation",
    "cancer",
    "mosaicism",
    "gene",
    "gene-expression",
    "virus",
    "bacteria",
    "metrological-traceability",
    "genome-editing",
    "dna-sequencing",
    "measurement-uncertainty"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}