compute_CCF: Compute CCF values.

In caravagnalab/CNAqc: CNAqc - Copy Number Analysis quality check

Compute CCF values.

Description

With this function, CNAqc can compute a CCF per mutation upon “phasing” the multiplicity for every input mutation. Phasing is the task of computing the number of copies of a mutation mapping in a certain copy number segment; this task is a difficult, and can lead to erroneous CCF estimates.

CNAqc computes CCFs for simple clonal CNA segments, offering two algorithms to phase mutations directly from VAFs.

* Entropy based method. The entropy-based approach will flag mutations for which we cannot phase multiplicity by VAFs with certainty; the CCFs of these mutations should be manually controlled and, unless necessary, discarded. To this aid, a QC pass is assigned with less than a certain percentage of mutations have uncertain CCFs. The model uses the entropy of a VAF mixture with two Binomial distributions to detect mutations happened before, and after aneuploidy. Assigning multiplicities is difficult at the crossing of the two densities where mutations could have multiplicity 1 or 2. If mistaken, these mutations can determine aritficial peaks in the CCF distribution and compromise downstream subclonal deconvolution.

* Hard-cut based method. A method is available to compute CCFs regardless of the entropy. From the 2-class Binomial mixture, CNAqc uses the means of the Binomial parameters to determine a hard split of the data. Since there are no NA assignments, the computation is always scored PASS for QC purposes; for this reason this computation is more “rough” than the one based on entropy.

Like for other analyses This function creates a field 'CCF_estimates' inside the returned object which contains the estimated CCFs.

Usage

compute_CCF(
  x,
  karyotypes = c("1:0", "1:1", "2:0", "2:1", "2:2"),
  muts_per_karyotype = 25,
  cutoff_QC_PASS = 0.1,
  method = "ENTROPY"
)

Arguments

x	A CNAqc object.
karyotypes	The karyotypes to use, this package supports only clonal simple CNAs.
muts_per_karyotype	Minimum number of mutations that are required to be mapped to a karyotype in order to compute CCF values (default 25).
cutoff_QC_PASS	For the entropy-based method, percentage of mutations that can be not-assigned (NA) in a karyotype. If the karyotype has more than cutoff_QC_PASS percentage of non-assigned mutations, then the overall set of CCFs is failed for the karyotype.
method	Either "ENTROPY" (default) or "ROUGH", to reflect the two different algorithms to compute CCF.

Value

A CNAqc object with CCF values.

See Also

Getters function CCF and plot_CCF.

Examples

data('example_dataset_CNAqc')
x = init(mutations = example_dataset_CNAqc$mutations, cna =example_dataset_CNAqc$cna, purity = example_dataset_CNAqc$purity)

x = compute_CCF(x, karyotypes = c('1:0', '1:1', '2:1', '2:0', '2:2'))
print(x)

# Extract the values with these other functions
CCF(x)
plot_CCF(x)

caravagnalab/CNAqc documentation built on Oct. 31, 2024, 3:54 a.m.