Cell-Free DNA (cfDNA) Sequencing

The MBCF cfDNA sequencing service provides a fully optimized, end-to-end workflow for targeted next-generation sequencing of cell-free DNA from patient plasma. The method is specifically engineered for low-input samples and incorporates duplex sequencing error suppression, enabling reliable detection of somatic variants at low variant allele frequencies (VAF).

This service supports custom-designed targeted capture panels. Because panel size and project-specific sensitivity targets drive sequencing depth requirements, all projects are quoted individually. 

Please contact us to discuss your project before submitting samples.

The workflow was developed and validated in-house at MBCF and consists of four core stages:

1. Adapter ligation — cfDNA fragments are ligated to adapters containing dual unique molecular identifiers (UMIs), which label both strands of each original DNA molecule. We use the IDT xGen cfDNA & FFPE DNA Library Preparation Kit, selected for its high ligation efficiency with low-input and fragmented DNA.

    

2. PCR amplification with icon96 — Library amplification is monitored in real time via fluorescence using the icon96 instrument. Rather than running a fixed number of PCR cycles, amplification is terminated automatically when each sample reaches a defined fluorescence threshold. This sample-specific cycle control compensates for variable input amounts and ligation efficiency resulting in normalized library yields.

    

3. Hybrid capture — Libraries are pooled and hybridized to custom-designed biotinylated probes targeting regions of interest. Probe design, hybridization conditions, and wash stringency have been co-optimized to achieve on-target sequencing rates above 90%.

    

4. Sequencing and duplex consensus calling — Sequencing is performed on the Illumina NovaSeq X Plus. After sequencing, reads from each strand of a UMI-tagged fragment are independently collapsed into single-strand consensus sequences. Complementary strand pairs are then matched to generate duplex consensus reads. Only variants present on both strands of an original molecule are called as true variants, reducing the effective sequencing error rate from approximately 10⁻³–10⁻⁴ (standard Illumina) to approximately 10⁻⁶–10⁻⁷ — a 100–1,000-fold improvement.

• Minimal residual disease (MRD) detection and monitoring
• Tumor evolution and clonal dynamics
• Treatment response monitoring via serial liquid biopsy
• Low-frequency somatic variant detection (SNVs, indels)
• Copy number variation (CNV) analysis
• Multi-gene custom targeted panels (~100 kb–1 Mb range)

Sensitivity in cfDNA sequencing is fundamentally limited by the number of input DNA molecules. More input material produces more duplex consensus reads, which directly increases the ability to detect low-frequency variants. The relationship between input amount and achievable sensitivity should be considered carefully when designing a project.

As a general guide based on our internal validation data:

• 1–10 ng input: Suitable for moderate-sensitivity applications. In validation experiments using reference standards, reliable variant detection was demonstrated at VAFs of approximately 0.5% (1 ng input) to 0.25% (10 ng input). These figures are based on controlled conditions and may not fully reflect performance with patient-derived samples of variable quality.
• Higher input (>10 ng, ideally >50 ng): Required for ultra-sensitive applications targeting VAFs below 0.1–0.25%. The practical lower detection limit for any given project depends on input amount, panel size, and achievable sequencing depth. Projects with ultra-low VAF targets should plan for higher input material where possible.

We will work with you during project planning to estimate the sequencing depth and input requirements needed to meet your sensitivity goals. Please note that these estimates will be conservative, and actual performance may vary depending on sample quality and tumor biology.