Calculate Cancer Cell Fraction Of Polyclonal

Estimate cancer cell fraction (CCF) from variant allele frequency, purity, copy number, and subclone-weighted multiplicity in a polyclonal tumor model.

Model used: CCF = VAF × (purity × CN_tumor + 2 × (1 – purity)) / (purity × m_effective), where m_effective is the subclone-weighted mutated copy count.

How to Calculate Cancer Cell Fraction in Polyclonal Tumors: Practical Expert Guide

Cancer cell fraction (CCF) is one of the most useful quantitative measures in modern cancer genomics. It estimates the proportion of cancer cells in a sample that carry a specific mutation. In simple terms, if a mutation has a CCF of 1.0, it is present in essentially all cancer cells (often called clonal). If it has a CCF of 0.35, then only about 35% of cancer cells carry that mutation (subclonal). In polyclonal tumors, this matters even more because different subclones coexist, each with distinct evolutionary histories and potentially different therapeutic vulnerabilities.

The challenge is that the observed variant allele frequency (VAF) in sequencing data is not the same thing as CCF. VAF is diluted by normal-cell contamination, affected by local copy number changes, and influenced by mutation multiplicity. A VAF of 0.2 can represent very different biological scenarios depending on whether purity is 30% or 90%, and whether the locus has copy-number gain or loss. That is why CCF calculations are essential in high-quality genomic interpretation.

Why Polyclonal Context Changes the Interpretation

In monoclonal assumptions, analysts often use a single mutation copy number (multiplicity) and a single clonal model. In polyclonal disease, however, one mutation may be carried by multiple subclones with different copy-number states. For example, one branch of the tumor phylogeny may carry the mutation on one copy, while another branch after duplication may carry two mutated copies. A weighted multiplicity is therefore a practical way to estimate CCF when you have mixed subclone architecture.

• Purity effect: lower tumor purity lowers observed VAF even when mutation is truly clonal.
• Copy number effect: CN amplification can reduce or increase observed VAF depending on mutated-copy dosage.
• Polyclonal effect: mixed subclones can produce the same VAF as a single clone, but imply different biology.
• Clinical effect: subclonal resistance mutations can be masked unless CCF is modeled correctly.

Core Formula Used in This Calculator

This calculator uses a standard purity and copy-number corrected CCF equation:

CCF = VAF × (p × CN_tumor + 2 × (1 – p)) / (p × m_effective)

Where p is tumor purity (0-1), and m_effective is the weighted mutated-copy number across subclones.

For the polyclonal extension, the effective multiplicity is: m_effective = w1 × m1 + w2 × m2, where w1 and w2 are normalized subclone shares and m1/m2 are mutated copies in each subclone. This is a practical approximation often used for quick interpretation before full Bayesian phylogenetic modeling.

Step-by-Step Workflow for Accurate CCF Estimation

1. Collect a high-confidence VAF from a well-filtered somatic callset.
2. Convert purity from percent to proportion (example: 70% to 0.70).
3. Use local total copy number from your segmentation or allele-specific copy-number pipeline.
4. Assign plausible mutated-copy multiplicities per major subclone.
5. Estimate relative subclone shares from clustering or phylogeny output.
6. Calculate weighted multiplicity and then compute CCF.
7. Interpret values above 1.0 as model tension, often due to underestimated multiplicity, purity error, or read count variance.

Reference Population Statistics for Context

When estimating CCF, it helps to understand cohort-level baselines. Large cancer genomics efforts provide key context for heterogeneity, purity, and clonality assumptions.

Program / Study	Scale	Relevant Statistic	Why It Matters for CCF
TCGA Pan-Cancer Atlas	33 tumor types, ~11,000 tumors	Large cross-cancer evidence of broad genomic heterogeneity and copy-number diversity	Supports routine purity and copy-number correction before CCF interpretation
PCAWG (ICGC/TCGA Whole Genomes)	2,658 whole cancer genomes	Substantial subclonal architecture observed across multiple tumor entities	Confirms that polyclonal modeling is often necessary
NCI SEER Program	Population-level US cancer surveillance	Millions of cases used for robust real-world oncology epidemiology	Reminds analysts that genomic findings must map to clinical population diversity

Authoritative sources: National Cancer Institute TCGA, NHGRI Cancer Genome Atlas resources, and SEER (NCI).

Scenario Comparison: Same VAF, Different Biology

The table below demonstrates how one observed VAF can map to very different inferred CCF values. These examples are calculated with the same formula used by the tool above and illustrate why simplistic VAF thresholds can be misleading in polyclonal disease.

Observed VAF	Purity	Total CN	Weighted Mutated Copies (meffective)	Estimated CCF	Interpretation
0.20	0.80	2.0	1.0	0.50	Likely subclonal
0.20	0.50	2.0	1.0	0.80	Could be near-clonal with normal contamination
0.20	0.70	4.0	1.0	1.03	Model suggests clonal or underestimated multiplicity
0.20	0.70	4.0	2.0	0.51	Subclonal once copy dosage is accounted for

Best Practices Before You Trust a CCF Number

• Check depth and strand balance: low-depth calls can inflate uncertainty dramatically.
• Use robust purity estimates: histology-only purity can diverge from molecular purity.
• Prefer allele-specific CN calls: total CN alone may hide LOH patterns.
• Cluster mutations jointly: single-variant CCF is less stable than cluster-level CCF.
• Quantify uncertainty: confidence intervals are critical for treatment-level decisions.

Interpreting Output in a Clinical-Research Pipeline

A practical interpretation framework is to treat CCF as a continuous measure, then map ranges to decision language:

• CCF ≥ 0.9: likely truncal or near-truncal event, especially if reproducible across regions/timepoints.
• CCF 0.3 to 0.9: mixed or branch-level event, often biologically meaningful in progression.
• CCF < 0.3: minor subclone candidate, sensitive to technical noise and sampling effects.

Keep in mind that longitudinal liquid biopsy and spatial sampling can reveal shifts in clonal fractions over time. A mutation that appears low-CCF at diagnosis may become dominant after treatment selection pressure. This is exactly why polyclonal-aware CCF analysis is increasingly integrated into translational pipelines, MRD research, and resistance monitoring.

Common Pitfalls in Polyclonal CCF Estimation

1. Assuming diploid CN=2 everywhere: this can severely bias CCF upward or downward.
2. Ignoring multiplicity: one mutated copy versus two makes a major difference.
3. Not normalizing subclone shares: percentages must be converted to weights summing to 1.
4. Treating CCF>1 as impossible error only: often it is a useful diagnostic signal of model misspecification.
5. Overinterpreting single mutations: robust clonality calls usually rely on mutation clusters.

Advanced Notes for Expert Users

If you need publication-grade inference, move from point estimation to probabilistic frameworks (for example, beta-binomial read count likelihoods combined with posterior distributions for purity and copy-number uncertainty). In those workflows, CCF becomes a posterior distribution rather than a single value. Still, this calculator remains valuable for rapid sensitivity analysis, tumor board preparation, and QC checks when integrating variant calls, purity estimates, and structural context.

You can also use this page to test hypotheses: adjust subclone shares and multiplicities, then inspect how CCF changes while VAF remains fixed. This is one of the fastest ways to teach teams why VAF is not a direct clonality measure in polyclonal tumors.

Bottom Line

To calculate cancer cell fraction in a polyclonal tumor correctly, you must adjust for purity, local copy number, and mutation multiplicity across subclones. The result is more biologically meaningful than raw VAF and can materially improve interpretation of clonal architecture, evolution, and potential therapeutic relevance. Use the calculator above as a practical front-end tool, then validate key findings in your full genomic pipeline.