What is Cell-Free DNA?

New to liquid biopsy? This page explains the biological foundation of cfDNA analysis. Already familiar? Skip to Getting Started.

The Liquid Biopsy Revolution

When cells in your body die—whether from normal turnover, injury, or disease—they release tiny pieces of their DNA into the bloodstream. These fragments, called cell-free DNA (cfDNA), circulate briefly before being cleared by the liver and kidneys.

Here's the breakthrough: cfDNA carries information about its source tissue.

For cancer patients, this means tumor DNA fragments (called ctDNA - circulating tumor DNA) mix with healthy cfDNA in the blood. By analyzing a simple blood draw, we can:

Detect cancer without a tissue biopsy
Monitor treatment response in real-time
Catch cancer recurrence earlier than imaging
Identify the tissue of origin for cancers of unknown primary

     ┌─────────────────────────────────────────────────────┐
     │                    Blood Sample                     │
     │  ┌───────────────────────────────────────────────┐  │
     │  │  cfDNA Pool                                   │  │
     │  │                                               │  │
     │  │  ~~~~  Normal cell DNA (~95-99%)             │  │
     │  │  ▓▓▓▓  Tumor DNA (ctDNA, ~1-5%)              │  │
     │  │                                               │  │
     │  └───────────────────────────────────────────────┘  │
     └─────────────────────────────────────────────────────┘

Why Fragment Sizes Matter

This is where it gets interesting. Cancer cells don't just release different DNA sequences—they release differently-sized fragments.

The Nucleosome Connection

DNA in your cells isn't floating freely—it's wrapped around protein spools called nucleosomes. Each nucleosome protects about 147 base pairs (bp) of DNA. When cells die, enzymes cut the DNA between nucleosomes, creating fragments.

                          Nucleosome (~147bp protected)
                          ┌──────────────────────┐
     ═══════════════════════════════════════════════════════
                 ▲                            ▲
                 │                            │
              Cut here                     Cut here
              (linker DNA)                (linker DNA)

The result: Most cfDNA fragments are ~166bp long (147bp nucleosome + ~20bp linker DNA).

Cancer Changes the Pattern

Tumor cells have abnormal chromatin (DNA packaging). This leads to:

Characteristic	Healthy cfDNA	Tumor cfDNA (ctDNA)
Dominant size	~166bp	~145bp (shorter!)
Size distribution	Sharp peak	Broader, left-shifted
10bp periodicity	Strong (DNA helix twist)	Often disrupted
Nucleosome pattern	Regular spacing	Irregular

Key insight: By measuring the ratio of short fragments (tumor-enriched) to long fragments (healthy-enriched), we can estimate tumor burden.

What Krewlyzer Extracts

Krewlyzer analyzes cfDNA sequencing data to extract fragmentomics features—numerical signatures that capture these biological patterns.

The Feature Categories

flowchart TB
    BAM[Blood Sample BAM] --> KREW[Krewlyzer]

    KREW --> SIZE[Fragment Size Features]
    KREW --> MOTIF[Cutting Pattern Features]
    KREW --> NUC[Nucleosome Features]
    KREW --> MUT[Mutation Features]

    SIZE --> FSC[FSC: Coverage by size]
    SIZE --> FSR[FSR: Short/Long ratio]
    SIZE --> FSD[FSD: Distribution per arm]

    MOTIF --> EDM[End Motifs: 4-mer frequencies]
    MOTIF --> MDS[MDS: Diversity score]

    NUC --> WPS[WPS: Protection scores]
    NUC --> OCF[OCF: Orientation patterns]

    MUT --> MFSD[mFSD: Mutant vs wild-type]

Use mouse to pan and zoom

What Each Feature Tells You

Feature	What It Measures	Clinical Application
FSR	Short vs long fragment ratio	Tumor burden estimation
FSD	Size distribution per chromosome arm	Copy number changes, aneuploidy
WPS	Nucleosome protection patterns	Tissue of origin, gene regulation
OCF	Fragment orientation at open chromatin	Tissue-specific cfDNA detection
MDS	Diversity of fragment end sequences	Chromatin accessibility changes
mFSD	Fragment sizes at known mutations	MRD monitoring, variant tracking

A Real-World Example

Let's walk through what happens when you analyze a cancer patient's blood:

1. Input: Sequenced Blood Sample

You have a BAM file from sequencing cfDNA extracted from a blood draw.

2. Run Krewlyzer

krewlyzer run-all -i patient_blood.bam -r hg19.fa -o results/

3. Examine Key Outputs

FSR (Fragment Size Ratio):

region              core_short_count  long_count  core_short_long_ratio
chr1:0-5000000      12,450           8,200       1.52
chr1:5000000-10M    11,890           7,950       1.50
...

A core_short_long_ratio of 1.5 (vs. ~0.9 in healthy samples) suggests elevated tumor burden.

FSD (Fragment Size Distribution): The size histogram shows a left-shifted peak at ~148bp instead of the healthy ~166bp.

WPS (Windowed Protection Score): Disrupted periodicity at ~190bp suggests abnormal nucleosome spacing—a hallmark of cancer.

Who Uses Krewlyzer?

User Type	Use Case
Cancer researchers	Studying tumor evolution, treatment resistance
Clinical lab developers	Building liquid biopsy diagnostic assays
Bioinformaticians	Creating ML models for early cancer detection
Pharmaceutical teams	Monitoring drug response in clinical trials
Translational scientists	Connecting fragmentomics to biology

Next Steps

Ready to start analyzing your data?

Installation - Get Krewlyzer running
Getting Started - Your first analysis in 5 minutes
Glossary - Terminology reference
Features Overview - Detailed feature documentation