get.perc: Computes percentiles for observations in the core set

In rxmenezes/rscreenorm: Normalization of Multiple Genetic Screens Data

Description

This is an internal function that computes percentiles for the data in the core set. This function is called by both get.rscreenorm and get.qnorm. It is to be used internally, so has no functionality other than what is described.

Usage

get.perc(data_rscreen, inv_set, n_perc = 1000)

Arguments

data_rscreen	an object of class 'rscreen.object', most likely corresponding to lethality scores produced by get.leth.scores. It is also possible to normalize re-scaled screen viability data, such as yielded by read.screen.data, or by combine.screens, although these tend to be less comparable between cell lines and replicates than the scores.
inv_set	a logical matrix with the same dimensions as the 'data_only' slot in data_rscreen, indicating the observations that belong to the core set. This is typically the output of sel.scores.inv.set.
n_perc	number of percentiles to be used to yield a representation of each core set. The default value is 1000, which is appropriate for pooled, whole genome screens. For small, hit-picking screens a smaller number may be more appropriate. The user may check the core set sizes to decide if in doubt. Note that the value of n_perc should not be larger than that of any core set.

Value

a matrix with as many rows as n_perc and as many columns as in the data_only slot of data_rscreen, containing the percentiles for the core sets, per replicate.

See Also

get.rscreenorm which yields core sets and quantile-normalizes data at once, codeget.qnorm which quantile-normalizes data for a given core set, get.inv.set for obtaining core sets for all replicates in the dataset and get.leth.scores to compute lethality scores.

rxmenezes/rscreenorm documentation built on May 15, 2019, 1:19 p.m.