adaCCP: Confirmatory cutpoint calculation

In schaarschmidt/mixADA: Normalization, mixture models and screening cutpoints for anti-drug-antibody reactions

Description

For internal use in the mixADA GUI: Compute confirmatory cutoiunts (CCP) as percentiles of ratios or percent inhibition of soiked and unspiked samples.

Usage

adaCCP(fitk2, rdat, ccplevel = 0.99, resp,
 nrdefinition = c("modelclass"), comparison = c("percinhib", "ratio"),
 aggfun = c("mean", "median"), runsmodel)

Arguments

fitk2	the object obtaiend from fitting a mixed effects mixture models using function xxx
rdat	the data set obtaiend from the normalization step, using function xxx
ccplevel	a single numeric value, ]0,1[, the level of the percentile to be computed
resp	a single character strinbg, providing the column name of the response variable in the data
nrdefinition
comparison	a single character string, specifying the type of comparison between the spiked an unspiked samples of the same subject, with each run
aggfun	character string naming the function to be used for aggregating repeated observations of the same subject within a technical replication
runsmodel	character string naming the column in the data set whcih summarized the technical replications

Details

Repeated observations of each spiked an unspiked sample within each level of runsmodel are summarized by aggfun. The resulting summary statistics are merged by sampleID and runsmodel For each of resulting pairs of spiked unspiked summary satistics for a given sampleID and runsmodel, spiked/unspiked or 100*(1 - spiked/unspiked) is computed if comparison=="ratio" or comparison=="percinhib", respectively. For the resulting values, two-sided percentile intervals with joint level ccplevel are computed for all sampleIDs, or only for the subset specified as non-reponder in fitk2.

Value

A list, use

DATCOMP	a data.frame containing the chosen summary stats for spiked as well as unspiked samples within each run
ynam	character string naming the statistic used for comparing spiked and unspiked samples
limtab	a data.frame with estimated percentiles/ limits to be used as CCP
comparison	character string for labelling the output of the GUI
infoccpmeasure	character string for labelling the output of the GUI
limitexplanation	character string for labelling the output of the GUI

schaarschmidt/mixADA documentation built on May 29, 2019, 3:25 p.m.