ROC.FCS {rflowcyt}R Documentation

ROC (Receiver Operating Characteristic) Curve: Percentage Positives for Flow Cytometry data

Description

This function plots an ROC curve based on cutoff values from the observed combined dataset of hivpos and hivneg, which both are vectors of patient-specific percentage positives based on the 99.9th percentile of the corresponding control sample distribution. The output contains the sensitivities, 1-specificity,and the observed dataset, cutoff values.

Usage

ROC.FCS(hivpos, hivneg, lineopt = 1, colopt = 1, overlay = FALSE)

Arguments

hivpos numerical vector of percentage positives for the HIV positive individuals/samples for a given condition
hivneg numerical vector of the percentage positives for the HIV negative individuals/samples for a given condition
lineopt numerical value for the lty option of the plot (line type)
colopt numerical value for the col option of the plot (color type)
overlay Boolean expression as to whether or not the plot is an overlay

Details

See 'PerPosROC' in the 'rfcdorig' package for a description of the input data and how percentage positives are defined.

The ROC curve in the example demonstrates that there is higher predictive ability of using the GAG stimulated samples rather than the PolA or PolB stimulated samples.

Value

Let T be the the percentage positives, c be a given value in c.obs, and HIV+ defined as among HIV positive individuals, and HIV- defined as among HIV negative individuals.

sensitivity numerical vector of the sensitivity=P(T>c | HIV+) calculated corresponding to a given cut-off in c.obs
spec.complement numerical vector of 1-specificity= P(T>c | HIV -)corresponding to a given cut-off in c.obs
c.obs a numerical vector of the cutoffs which were taken to be the values of the observations (the values of the percentage positives of both the HIV positive and HIV negative data)

Author(s)

A.J. Rossini and J.Y. Wan

References

Zoe Moodie and Mario Roederer

See Also

PercentPos.FCS, data 'PerPosROC' in 'rfcdorig' package, percentile.FCS

Examples


if (require(rfcdmin)){

data(PerPosROCmin)

#plotting the gag stimulated 100* percent positives
if (interactive()==TRUE){
GAG<-ROC.FCS(hivpos.gag, hivneg.gag)
#plotting the pola stimulated 100* percent positives
POLA<-ROC.FCS(hivpos.pola, hivneg.pola, lineopt=2, colopt=2, overlay=TRUE)
#plotting the polb stimulated 100* percent positives
POLB<-ROC.FCS(hivpos.polb, hivneg.polb, lineopt=4, colopt=3, overlay=TRUE)
legend(0.7, 0.7, c("gag", "polA", "polB"), col = c(1,2,3), lty=c(1,2,4))
}

}

[Package rflowcyt version 1.14.0 Index]