cfDNA and ctDNA Sample Prep for Liquid Biopsy: A Guide
Team Dynamic Matrices | 25 April 2026
Cell-free DNA is a difficult analyte for three reasons at once: it is short, it is scarce, and it sits in a background that actively gets worse from the moment blood leaves the arm. The tumour or fetal fraction you want is often under one percent of the total. Every step from the collection tube to the elution either preserves that fraction or dilutes it. This guide is what we tell partners setting up a liquid biopsy workflow.
What cfDNA is, physically
Plasma cfDNA is released mainly by apoptosis and is protected from nucleases by nucleosomes. The result is a fragment distribution with a sharp mononucleosomal peak at about 166 bp, a smaller dinucleosomal peak near 330 bp, and a 10 bp periodicity from nuclease cutting around the histone core (Snyder et al., Cell, 2016). Tumour-derived fragments are, on average, shorter than the background, peaking around 130 to 150 bp (Mouliere et al., Science Translational Medicine, 2018), which is both a QC clue and a basis for size selection. Typical yield is 1 to 10 ng per millilitre of plasma in healthy donors, higher in cancer, inflammation and pregnancy.
Two things follow. First, because the target is already fragmented, any further shearing or nuclease exposure during processing shifts the distribution away from the biologically informative peaks. Second, the background is alive: leukocytes in the tube lyse within hours and release high-molecular-weight genomic DNA that dilutes the tumour fraction (Lui et al., Clinical Chemistry, 2002).
Collection and plasma preparation
Tubes. K2-EDTA tubes are fine if plasma is separated within 2 hours, ideally within 1. For anything longer, use cell-stabilising tubes (Streck cfDNA BCT, PAXgene ccfDNA, Roche cfDNA) which hold leukocytes intact for several days at room temperature. Do not use heparin; it inhibits downstream enzymes.
Double spin. First spin at 800 to 1,600 g for 10 minutes to pellet cells; transfer plasma without disturbing the buffy coat; second spin at 16,000 g for 10 minutes to remove platelets and debris. Skipping the second spin leaves platelets that lyse on freezing and release DNA. Aliquot and freeze at minus 80 °C; avoid repeat freeze-thaw.
Record tube type, time from draw to spin, and haemolysis grade. These three explain more inter-sample variation than the extraction kit.
Extraction for short fragments
Use a chemistry validated for cfDNA (silica-membrane kits with cfDNA-specific binding conditions, or bead-based kits designed for short fragments). Kits adapted from genomic DNA protocols lose the short end of the distribution. Elute in a small volume; cfDNA concentrations are low and dilution hurts.
Before anything downstream, check three numbers:
- Yield by fluorometry (Qubit dsDNA HS), never absorbance, which cannot see nanogram quantities accurately.
- Fragment size distribution by capillary electrophoresis (Bioanalyzer HS or TapeStation cfDNA assay). You want a clean 166 bp peak.
- The ratio of the mononucleosomal peak to high-molecular-weight signal. A shoulder above 1,000 bp means leukocyte gDNA contamination, and no downstream capture will fully remove it.
Selective capture: where the fraction gets lost
For ctDNA panels and NIPT, the workflow usually captures a target region set, or depletes background, before sequencing. This is where a sub-one-percent signal is most easily lost:
- Off-target co-capture. Genomic fragments sharing partial homology with panel probes come along, particularly with bead-based hybrid capture where surface adsorption adds non-specific background.
- Harsh elution. Releasing DNA from streptavidin needs heat or denaturant. Already short fragments do not survive that well; single-stranded release also complicates downstream ligation.
- Surface losses. At nanogram input, adsorption to beads and tubes takes a disproportionate share of a scarce sample.
We designed LASSO capture to run under mild, native conditions with no solid surface: catcher strands on a soluble polymer hybridise to target fragments, the network condenses into a pellet, and release is by strand displacement in the same buffer. In our hands this keeps recovery high and off-target low at the input levels cfDNA workflows operate at. We are currently running long-fragment gDNA capture feasibility work with a clinical lab partner using the same chemistry.
QC gates, not spot checks
- Plasma processed within the tube's stabilisation window
- One freeze-thaw cycle
- Fluorometric quantification
- Size distribution checked on every sample before library prep
- gDNA contamination ratio as a pass/fail gate, on every sample
- Spike-in control at extraction to measure recovery
Treat these as standing gates. The failures they catch are invisible in the sequencing report until the variant you needed is missing.
References
- Snyder MW, Kircher M, Hill AJ, Daza RM, Shendure J. Cell-free DNA comprises an in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164, 57–68 (2016).
- Mouliere F et al. Enhanced detection of circulating tumor DNA by fragment size analysis. Sci Transl Med 10, eaat4921 (2018).
- Lui YYN et al. Predominant hematopoietic origin of cell-free DNA in plasma and serum after sex-mismatched bone marrow transplantation. Clin Chem 48, 421–427 (2002).
- Meddeb R, Pisareva E, Thierry AR. Guidelines for the preanalytical conditions for analyzing circulating cell-free DNA. Clin Chem 65, 623–633 (2019).
- Dynamic Matrices. LASSO: sequence-selective biomolecule isolation by programmable polymer phase separation. Angew Chem Int Ed (2025); patent WO 2025/196123A1.
FAQs
A dominant peak around 166 bp, with tumour-enriched samples shifted toward 130 to 150 bp. Peaks above 1,000 bp indicate leukocyte gDNA.
They extend the window to days. Same-day or next-day processing is still best practice when logistics allow.
In silico or physical selection for 90 to 150 bp fragments can roughly double tumour fraction (Mouliere 2018), at the cost of total yield. Worth it for low-fraction monitoring; unnecessary for high-burden samples.
Cell-free DNA (cfDNA) is all the DNA circulating in plasma, released mainly by apoptosis and largely from blood cells. Circulating tumour DNA (ctDNA) is the tumour-derived fraction of it, often well under one percent of the total, which is why collection, extraction and capture choices matter so much.
ctDNA is extracted together with all other cfDNA and identified afterwards by tumour-specific variants. Use a chemistry validated for cfDNA, with silica-membrane or bead-based kits designed for short fragments; process plasma within the tube's stabilisation window; elute in a small volume; and check yield by fluorometry and fragment size by capillary electrophoresis. Extraction does not separate tumour DNA from other cfDNA, so enrichment comes from size selection or capture.
Three numbers: yield by fluorometry (Qubit dsDNA HS), fragment size distribution by capillary electrophoresis with a clean peak near 166 bp, and the ratio of that peak to high-molecular-weight signal. A shoulder above 1,000 bp means leukocyte genomic DNA contamination that no downstream capture will fully remove.
Library prep for ctDNA uses chemistry designed for short, scarce fragments, usually with unique molecular identifiers to separate real variants from errors, followed by hybrid capture or amplicon enrichment. Input sets the limit of detection. See ctDNA library prep and enrichment.