01-cnvPrior: Class "CNVPrior"

In DeepCNV: Exploiting Normal Contamination to Infer Copy Number from Deep Sequencing

Description

Represents the prior distribution used to infer copy number from targeted deep sequencing.

Usage

setCNVPrior(alpha = 1, beta = 1,
            pAbnormal=2/3, pGain = NULL, pLoss = pGain,
            pMutWrong = 0.01, pSnpWrong = 0.01, resn=300)
cnPrior(object)
## S4 method for signature 'CNVPrior,missing'
plot(x, lwd=2, ...)
## S4 method for signature 'CNVPrior'
summary(object, ...)

Arguments

alpha	Parameter for the beta prior distribution on the fraction of normal cells.
beta	Parameter for the beta prior distribution on the fraction of normal cells.
pAbnormal	Prior probability of an aberration; split equally between gain and loss.
pGain	Prior probability of a gain in copy number. Defaults to NULL, in which case it is set to half of pAbnormal.
pLoss	Prior probability of a loss of copy number. Defaults to NULL, in which case it is set to half of pAbnormal.
pMutWrong	Prior probability that something called a somatic mutation is actually a known single nucleotide polymorphism (SNP).
pSnpWrong	Prior probability that something called a SNP is actually a somatic mutation.
resn	The number of points at which to compute values of the beta distribution in order to compute the likelihood and posterior.
object	object of class CNVPrior.
x	object of class CNVPrior.
...	extra arguments for generic or plotting routines.
lwd	Line width parameter.

Details

The DeepCNV class is used to fit a Bayesian model to targeted sequencing data from one or a few genes in order to draw inferences about possible copy number changes. Basically, we assume that the observed data consists of a list of triples (K, N, V), one for each variant in a gene. Here K is the number of variant reads, N is the total number of reads, and V is the type of each variant (either a known SNP or a somatic mutation). We model (K, N) using a binomial distribution, where the 'success' parameter depends (in a deterministic way) on the unknown parameters of interest: the fraction ν of normal cells in the sample and the copy number state (Normal, Deleted, or Gained).

The prior distribution consists of a continuous (by default, Beta(α, β)) distribution on ν and a discrete distribution on the copy number state S. The discrete distribution can be defined either by setting pAbnormal, in which case the probability os divided equally between Deleted and Gained, or by setting both pGain and pLoss.

Note that the settings of pGain and pLoss, if present, override any setting of pAbnormal.

One complication arises from the variant type V, which can be observed with error. In some cases, observations of V are compared with a matched normal sample of DNA from the same subject, in which case the variant type is known with great confidence. In the absence of matched normal samples, however, the variant type may be inferred indirectly from an external database like dbSNP, in which case the confidence decreases. Our confidence in these calls is represented by the prior probabilities assigned in pMutWrong and pSnpWrong. For purposes of our analysis, V is a nuisance parameter that we mostly try to hide from the user. However, the full structure of the discrete prior is revealed when using the summary method to look at the prior.

Value

The setCNVPrior constructor returns a valid object of the CNVPrior class.

The cnPrior function returns a vector of length three containing the prior probabilities of the copy number states.

Slots

discrete:

Data frame containing the discrete part of the prior.

pGain:

Prior probability of a gain in copy number.

pLoss:

Prior probability of a loss of copy number.

pMutWrong

Prior probability that something called a somatic mutation is actually a known single nucleotide polymorphism (SNP).

pSnpWrong

Prior probability that something called a SNP is actually a somatic mutation.

alpha

Parameter for the beta prior distribution on the fraction of normal cells.

beta

Parameter for the beta prior distribution on the fraction of normal cells.

grid:

Realized grid of points at which the continuous prior is computed.

pdf:

Values of the prior distribution at the grid points.

Methods

plot(x, ...)

PLots the continuous part of the prior distribution.

summary(object, ...)

Writes out a table detailing the discrete part of the prior distribution.

Author(s)

Kevin R. Coombes [e-mail]

See Also

CNVPosterior

Examples

prior <- setCNVPrior(alpha=1.2, beta=4.8, pAbnormal=0.6)
cnPrior(prior)
summary(prior)
plot(prior)

DeepCNV documentation built on May 2, 2019, 5:23 p.m.