spkGNN {spkTools} | R Documentation |
Computes the number of genes one would need to consider to obtain a given number of truly positive genes if one considered genes in order of decreasing observed fold change.
spkGNN(n, n.expr, n.unexpr, AccuracySlope, AccuracySD, nullfc)
n |
the desired number of true positives |
n.expr |
the actual number of truly expressed genes |
n.unexpr |
the actual number of truly unexpressed genes |
AccuracySlope |
the signal detect slope from the spkSlope function |
AccuracySD |
the standard deviation of the signal detect slope from the spkAccSD function |
nullfc |
a vector of null fold changes from the spkBox function |
This function returns the expected number of genes one would have to consider to obtain N true positives under the given conditions.
Matthew N. McCall
data(affy) spkSlopeOut <- spkSlope(affy) spkBoxOut <- spkBox(affy, spkSlopeOut, fc=2) AccuracySlope <- round(spkSlopeOut$slope[-1], digits=2) AccuracySD <- round(spkAccSD(affy, spkSlopeOut), digits=2) spkGNN(n=25, n.expr=100, n.unexpr=10000, AccuracySlope[2], AccuracySD[2], spkBoxOut[[2]])