R: Calculate the number of true positives and false positives

getTpFp

R Documentation

Calculate the number of true positives and false positives

Description

Calculate the number of true positives and false positives among candidate breakpoints

Usage

getTpFp(candidates, trueBkp, tol, relax = -1)

Arguments

`candidates`	Breakpoints found by the methods
`trueBkp`	True breakpoints
`tol`	Tolerance on the position of candidate breakpoints called true
`relax`	Controls the way multiple breapoints within tolerance area are recorded. 1 count one true positive if there is at least one breakpoint within tolerance area 0 count one true positive only if there is exactly one breakpoint within tolerance area -1 count only one true positive if there is exactly one breakpoint within tolerance area; other breakpoints are counted as false positives

Value

A list with elements:

TP: The number of true positives
FP: The number of false positives

Author(s)

Morgane Pierre-Jean and Pierre Neuvial

Examples


## load known real copy number regions
affyDat <- acnr::loadCnRegionData(dataSet="GSE29172", tumorFraction=0.7)

## generate a synthetic CN profile
K <- 10
len <- 2e4
sim <- getCopyNumberDataByResampling(len, K, minLength=100, regData=affyDat)
datS <- sim$profile

## (group-)fused Lasso segmentation
res <- PSSeg(data=datS, K=2*K, method="GFLars", stat="c", profile=TRUE)

## results of the initial (group-)fused lasso segmentation
getTpFp(res$initBkp, sim$bkp, tol=10, relax=-1)
getTpFp(res$initBkp, sim$bkp, tol=10, relax=0)
getTpFp(res$initBkp, sim$bkp, tol=10, relax=1)
plotSeg(datS, breakpoints=list(sim$bkp, res$initBkp))

## results after pruning (group-)fused Lasso candidates by dynamic programming)
getTpFp(res$bestBkp, sim$bkp, tol=10, relax=-1)
getTpFp(res$bestBkp, sim$bkp, tol=10, relax=0)
getTpFp(res$bestBkp, sim$bkp, tol=10, relax=1)
plotSeg(datS, breakpoints=list(sim$bkp, res$bestBkp))