modify.conf: Modify the configuration (of calls) of the plrs model

In plrs: Piecewise Linear Regression Splines (PLRS) for the association between DNA copy number and gene expression

Description

This function changes the discrete copy number values for a given gene in order to force a minimum number of observations per state.

Usage

modify.conf(cghcall, min.obs = 3, discard = TRUE)

Arguments

cghcall	Vector of called values
min.obs	Minimum number of observations per state
discard	Logical. Whether discrete states with few observations should be discarded from analysis.

Details

Consider that the number of observations of a given state is lower than min.obs, then:

- if discard = FALSE, observations are not discarded and a rearrangement of called values is carried out as follows. The "normal" copy number state is taken as a reference. If the minimum number of observations is not obtained, "losses" will be merged to "normals", "gains" to "normals" and "amplifications" to "gains". Note that this modifies the configuration of the model. Thus, after fitting a model using plrs, original and modified data are stored in the resulting plrs-class object, respectively under slots data and mdata.

- if discard = TRUE, states for which the number of observations is lower than min.obs are discarded (replaced by NAs).

Value

val	Vector of new called values

Note

This function is implemented within function plrs and plrs.series.

Author(s)

Gwenael G.R. Leday [e-mail]

Examples

called <- sample(c(rep(-1,5),rep(0,15),rep(1,2),rep(2,1)))
table(called)
table(modify.conf(called, min.obs=3))

plrs documentation built on April 28, 2020, 6:09 p.m.