R/MultipleMockData.R
In mbertalan/iPsychCNV: iPsychCNV
##' MultipleMockData: Runs mock data multiple times.
##'
##' Specifically designed to handle noisy data from amplified DNA on phenylketonuria (PKU) cards. The function is a pipeline using many subfunctions.
##' @title MultipleMockData
##' @param NSamples: Unknown, default = 10.
##' @param NLoops: Unknown, default = 10.
##' @param Cores: Number of cores used, default = 30.
##' @param HMM: Unknown, default = Unknown.
##' @param Path2PennCNV: The path to pennCNV.
##' @return Return data in a data frame.
##' @author Marcelo Bertalan, Louise K. Hoeffding.
##' @source \url{http://biopsych.dk/iPsychCNV}
##' @export
##' @examples Unknown.
##'
MultipleMockData <- function(NSamples=100, NLoops=10, Cores=28, HMM="/media/NeoScreen/NeSc_home/share/Programs/penncnv/lib/hhall.hmm", Path2PennCNV="/media/NeoScreen/NeSc_home/share/Programs/penncnv/")
{
library(pROC)
Res <- sapply(1:NLoops, function(Loops)
{
# Creating Mock data
cat("Creating mock data: Loop ", Loops, "\n")
MockDataCNVs <- MockData(N=NSamples, Type="PKU", Cores=Cores)
MockHot <- subset(MockDataCNVs, ID %in% "MockSample_1.tab" & CN != 2)
MockHot <- as.data.frame(MockHot)
MockHot$Class <- "ROI"
# iPsychCNV prediction
cat("Running iPsychCNV: Loop ", Loops, "\n" )
iPsych.Pred <- iPsychCNV(PathRawData=".", MINNumSNPs=20, Cores=Cores, Pattern="^MockSample", MinLength=10, Skip=0, LCR=FALSE, Quantile=FALSE)
iPsych.Pred <- subset(iPsych.Pred, CN != 2)
# iPsychCNV + Hotspots + ReScanCNVs
#CNVs.Hotspots <- HotspotsCNV(df=iPsych.Pred, Freq=3, OverlapCutoff=0.8, Cores=Cores)
#iPsych.rescan <- ReScanCNVs(CNVs=MockHot, Cores=Cores, Pattern="^MockSample_*", Skip=0, hg="hg19", PathRawData=".", IndxPos=TRUE)
# PennCNV
cat("Running PennCNV: Loop ", Loops, "\n")
PennCNV.Pred <- RunPennCNV(PathRawData=".", Pattern="^MockSample.*", Cores=Cores, Skip=0, Normalization=FALSE, PFB="NO", HMM=HMM, Path2PennCNV=Path2PennCNV)
# PlotAllCNVs
NewName <- paste("iPsych.Pred.", Loops, ".png", sep="", collapse="")
PlotAllCNVs(df=iPsych.Pred, Name=NewName, Roi=MockHot)
NewName <- paste("PennCNV.Pred.", Loops, ".png", sep="", collapse="")
PlotAllCNVs(df=PennCNV.Pred, Name=NewName, Roi=MockHot)
# PennCNV Filter
#PennCNV.filter <- FilterFromCNVs(CNVs=PennCNV.Pred, PathRawData=".", MinNumSNPs=10, Source="PennCNV.Filter", Skip=0, Cores=Cores)
# Filter + Hotspots + ReScanCNVs
#tmp <- subset(PennCNV.filter, CN != 2)
#CNVs.Hotspots <- HotspotsCNV(df=tmp[,1:20], Freq=1, OverlapCutoff=0.8, Cores=Cores)
#PennCNV.rescan <- ReScanCNVs(CNVs=CNVs.Hotspots, Cores=Cores, Pattern="^MockSample_*", Skip=0, hg="hg19", PathRawData=".")
#PennCNV.rescan$ID <- PennCNV.rescan$SampleID
# Evaluating methods
cat("iPsychCNV evaluting methods: Loop ", Loops, "\n")
iPsychCNV.Eval <- EvaluateMockResults(MockDataCNVs, iPsych.Pred, Cores=Cores)
cat("PennCNV Evaluting methods: Loop ", Loops, "\n")
PennCNV.Eval <- EvaluateMockResults(MockDataCNVs, PennCNV.Pred, Cores=Cores)
#cat("ReiPsychCNV Evaluting methods: Loop ", Loops, "\n")
#ReiPsych.Eval <- EvaluateMockResults(MockDataCNVs, iPsych.rescan, Cores=Cores)
cat("Done evaluting methods: Loop ", Loops, "\n")
#Filter.Eval <- EvaluateMockResults(MockDataCNVs, PennCNV.filter, Cores=Cores)
#Rescan.Eval <- EvaluateMockResults(MockDataCNVs, PennCNV.rescan, Cores=Cores)
# ROC
iPsychCNV.AUC <- roc(iPsychCNV.Eval$CNV.Predicted, iPsychCNV.Eval$CNV.Present)
#ReiPsych.AUC <- roc(ReiPsych.Eval$CNV.Predicted, ReiPsych.Eval$CNV.Present)
PennCNV.AUC <- roc(PennCNV.Eval$CNV.Predicted, PennCNV.Eval$CNV.Present)
#Filter.AUC <- roc(Filter.Eval$CNV.Predicted, Filter.Eval$CNV.Present)$auc[1]
#Rescan.AUC <- roc(Rescan.Eval$CNV.Predicted, Rescan.Eval$CNV.Present)$auc[1]
# Coords
iPsychCNV.AUC.coords <- coords(iPsychCNV.AUC, "best", ret=c("specificity", "sensitivity", "accuracy", "tn", "fn", "tp", "fp", "npv", "ppv"))
#ReiPsych.AUC.coords <- coords(ReiPsych.AUC, "best", ret=c("specificity", "sensitivity", "accuracy", "tn", "fn", "fp", "npv", "ppv"))
PennCNV.AUC.coords <- coords(PennCNV.AUC, "best", ret=c("specificity", "sensitivity", "accuracy", "tn", "fn", "tp", "fp", "npv", "ppv"))
iPsychCNV.AUC.coords <- as.data.frame(t(iPsychCNV.AUC.coords))
#ReiPsych.AUC.coords <- as.data.frame(t(ReiPsych.AUC.coords))
PennCNV.AUC.coords <- as.data.frame(t(PennCNV.AUC.coords))
iPsychCNV.AUC.coords$auc <- iPsychCNV.AUC$auc[1]
#ReiPsych.AUC.coords$auc <- ReiPsych.AUC$auc[1]
PennCNV.AUC.coords$auc <- PennCNV.AUC$auc[1]
iPsychCNV.AUC.coords$source <- "iPsychCNV"
#ReiPsych.AUC.coords$source <- "Re-ScanCNVs"
PennCNV.AUC.coords$source <- "PennCNV"
iPsychCNV.AUC.coords$Loop <- Loops
PennCNV.AUC.coords$Loop <- Loops
ValuesDF <- rbind(iPsychCNV.AUC.coords, PennCNV.AUC.coords)
# When prediction fail.
## PennCNV
tmp <- subset(PennCNV.Eval, CN != 2)
tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
tmp6 <- tapply(tmp$True.Positive, as.factor(tmp$Heterozygosity), function(X){ sum(X)/length(X) })
df <- data.frame(AUC=PennCNV.AUC$auc[1], True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5), as.numeric(tmp6)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5)), rep("Heterozygosity", length(tmp6))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5), names(tmp6)), Source="PennCNV", Loop=Loops, stringsAsFactors=F)
## iPsychCNV
tmp <- subset(iPsychCNV.Eval, CN != 2)
tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
tmp6 <- tapply(tmp$True.Positive, as.factor(tmp$Heterozygosity), function(X){ sum(X)/length(X) })
#df2 <- data.frame(AUC=iPsychCNV.AUC,True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="iPsychCNV", Loop=Loops, stringsAsFactors=F)
df2 <- data.frame(AUC=iPsychCNV.AUC$auc[1], True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5), as.numeric(tmp6)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5)), rep("Heterozygosity", length(tmp6))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5), names(tmp6)), Source="iPsychCNV", Loop=Loops, stringsAsFactors=F)
## iPsychCNV + ReScan
#tmp <- subset(ReiPsych.Eval, CN != 2)
#tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
#tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
#tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
#tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
#tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
#tmp6 <- tapply(tmp$True.Positive, as.factor(tmp$Heterozygosity), function(X){ sum(X)/length(X) })
#df3 <- data.frame(AUC=ReiPsych.AUC,True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="iPsychCNV + ReScan", Loop=Loops, stringsAsFactors=F)
#df3 <- data.frame(AUC=ReiPsych.AUC, True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5), as.numeric(tmp6)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5)), rep("Heterozygosity", length(tmp6))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5), names(tmp6)), Source="iPsychCNV + ReScan", Loop=Loops, stringsAsFactors=F)
## PennCNV + Filter
#tmp <- subset(Filter.Eval, CN != 2)
#tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
#tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
#tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
#tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
#tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
#df3 <- data.frame(AUC=Filter.AUC, True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="Filter", Loop=Loops, stringsAsFactors=F)
## PennCNV + Filter + Hotspots + ReScan
#tmp <- subset(Rescan.Eval, CN != 2)
#tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
#tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
#tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
#tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
#tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
#df4 <- data.frame(AUC=Rescan.AUC, True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="ReScan", Loop=Loops, stringsAsFactors=F)
df5 <- rbind(df, df2)
#df5 <- rbind(df, df2, df3)
#df5 <- rbind(df, df2, df3, df4)
system("rm -f MockSample_*")
return(ValuesDF)
})
Res <- MatrixOrList2df(Res)
#Res$Values <- as.numeric(Res$Values)
return(Res)
}
# Res <- MultipleMockData(NSamples=30, NLoops=30)
# ggplot(Res, aes(x=as.factor(Values), y=True.positive)) + geom_boxplot(aes(fill=as.factor(Source)), outlier.shape=NA) + facet_wrap(~ By, scales = "free") + scale_fill_brewer(palette="Set1")
mbertalan/iPsychCNV documentation built on May 22, 2019, 12:19 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
##' MultipleMockData: Runs mock data multiple times.
##'
##' Specifically designed to handle noisy data from amplified DNA on phenylketonuria (PKU) cards. The function is a pipeline using many subfunctions.
##' @title MultipleMockData
##' @param NSamples: Unknown, default = 10.
##' @param NLoops: Unknown, default = 10.
##' @param Cores: Number of cores used, default = 30.
##' @param HMM: Unknown, default = Unknown.
##' @param Path2PennCNV: The path to pennCNV.
##' @return Return data in a data frame.
##' @author Marcelo Bertalan, Louise K. Hoeffding.
##' @source \url{http://biopsych.dk/iPsychCNV}
##' @export
##' @examples Unknown.
##'
MultipleMockData <- function(NSamples=100, NLoops=10, Cores=28, HMM="/media/NeoScreen/NeSc_home/share/Programs/penncnv/lib/hhall.hmm", Path2PennCNV="/media/NeoScreen/NeSc_home/share/Programs/penncnv/")
{
library(pROC)
Res <- sapply(1:NLoops, function(Loops)
{
# Creating Mock data
cat("Creating mock data: Loop ", Loops, "\n")
MockDataCNVs <- MockData(N=NSamples, Type="PKU", Cores=Cores)
MockHot <- subset(MockDataCNVs, ID %in% "MockSample_1.tab" & CN != 2)
MockHot <- as.data.frame(MockHot)
MockHot$Class <- "ROI"
# iPsychCNV prediction
cat("Running iPsychCNV: Loop ", Loops, "\n" )
iPsych.Pred <- iPsychCNV(PathRawData=".", MINNumSNPs=20, Cores=Cores, Pattern="^MockSample", MinLength=10, Skip=0, LCR=FALSE, Quantile=FALSE)
iPsych.Pred <- subset(iPsych.Pred, CN != 2)
# iPsychCNV + Hotspots + ReScanCNVs
#CNVs.Hotspots <- HotspotsCNV(df=iPsych.Pred, Freq=3, OverlapCutoff=0.8, Cores=Cores)
#iPsych.rescan <- ReScanCNVs(CNVs=MockHot, Cores=Cores, Pattern="^MockSample_*", Skip=0, hg="hg19", PathRawData=".", IndxPos=TRUE)
# PennCNV
cat("Running PennCNV: Loop ", Loops, "\n")
PennCNV.Pred <- RunPennCNV(PathRawData=".", Pattern="^MockSample.*", Cores=Cores, Skip=0, Normalization=FALSE, PFB="NO", HMM=HMM, Path2PennCNV=Path2PennCNV)
# PlotAllCNVs
NewName <- paste("iPsych.Pred.", Loops, ".png", sep="", collapse="")
PlotAllCNVs(df=iPsych.Pred, Name=NewName, Roi=MockHot)
NewName <- paste("PennCNV.Pred.", Loops, ".png", sep="", collapse="")
PlotAllCNVs(df=PennCNV.Pred, Name=NewName, Roi=MockHot)
# PennCNV Filter
#PennCNV.filter <- FilterFromCNVs(CNVs=PennCNV.Pred, PathRawData=".", MinNumSNPs=10, Source="PennCNV.Filter", Skip=0, Cores=Cores)
# Filter + Hotspots + ReScanCNVs
#tmp <- subset(PennCNV.filter, CN != 2)
#CNVs.Hotspots <- HotspotsCNV(df=tmp[,1:20], Freq=1, OverlapCutoff=0.8, Cores=Cores)
#PennCNV.rescan <- ReScanCNVs(CNVs=CNVs.Hotspots, Cores=Cores, Pattern="^MockSample_*", Skip=0, hg="hg19", PathRawData=".")
#PennCNV.rescan$ID <- PennCNV.rescan$SampleID
# Evaluating methods
cat("iPsychCNV evaluting methods: Loop ", Loops, "\n")
iPsychCNV.Eval <- EvaluateMockResults(MockDataCNVs, iPsych.Pred, Cores=Cores)
cat("PennCNV Evaluting methods: Loop ", Loops, "\n")
PennCNV.Eval <- EvaluateMockResults(MockDataCNVs, PennCNV.Pred, Cores=Cores)
#cat("ReiPsychCNV Evaluting methods: Loop ", Loops, "\n")
#ReiPsych.Eval <- EvaluateMockResults(MockDataCNVs, iPsych.rescan, Cores=Cores)
cat("Done evaluting methods: Loop ", Loops, "\n")
#Filter.Eval <- EvaluateMockResults(MockDataCNVs, PennCNV.filter, Cores=Cores)
#Rescan.Eval <- EvaluateMockResults(MockDataCNVs, PennCNV.rescan, Cores=Cores)
# ROC
iPsychCNV.AUC <- roc(iPsychCNV.Eval$CNV.Predicted, iPsychCNV.Eval$CNV.Present)
#ReiPsych.AUC <- roc(ReiPsych.Eval$CNV.Predicted, ReiPsych.Eval$CNV.Present)
PennCNV.AUC <- roc(PennCNV.Eval$CNV.Predicted, PennCNV.Eval$CNV.Present)
#Filter.AUC <- roc(Filter.Eval$CNV.Predicted, Filter.Eval$CNV.Present)$auc[1]
#Rescan.AUC <- roc(Rescan.Eval$CNV.Predicted, Rescan.Eval$CNV.Present)$auc[1]
# Coords
iPsychCNV.AUC.coords <- coords(iPsychCNV.AUC, "best", ret=c("specificity", "sensitivity", "accuracy", "tn", "fn", "tp", "fp", "npv", "ppv"))
#ReiPsych.AUC.coords <- coords(ReiPsych.AUC, "best", ret=c("specificity", "sensitivity", "accuracy", "tn", "fn", "fp", "npv", "ppv"))
PennCNV.AUC.coords <- coords(PennCNV.AUC, "best", ret=c("specificity", "sensitivity", "accuracy", "tn", "fn", "tp", "fp", "npv", "ppv"))
iPsychCNV.AUC.coords <- as.data.frame(t(iPsychCNV.AUC.coords))
#ReiPsych.AUC.coords <- as.data.frame(t(ReiPsych.AUC.coords))
PennCNV.AUC.coords <- as.data.frame(t(PennCNV.AUC.coords))
iPsychCNV.AUC.coords$auc <- iPsychCNV.AUC$auc[1]
#ReiPsych.AUC.coords$auc <- ReiPsych.AUC$auc[1]
PennCNV.AUC.coords$auc <- PennCNV.AUC$auc[1]
iPsychCNV.AUC.coords$source <- "iPsychCNV"
#ReiPsych.AUC.coords$source <- "Re-ScanCNVs"
PennCNV.AUC.coords$source <- "PennCNV"
iPsychCNV.AUC.coords$Loop <- Loops
PennCNV.AUC.coords$Loop <- Loops
ValuesDF <- rbind(iPsychCNV.AUC.coords, PennCNV.AUC.coords)
# When prediction fail.
## PennCNV
tmp <- subset(PennCNV.Eval, CN != 2)
tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
tmp6 <- tapply(tmp$True.Positive, as.factor(tmp$Heterozygosity), function(X){ sum(X)/length(X) })
df <- data.frame(AUC=PennCNV.AUC$auc[1], True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5), as.numeric(tmp6)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5)), rep("Heterozygosity", length(tmp6))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5), names(tmp6)), Source="PennCNV", Loop=Loops, stringsAsFactors=F)
## iPsychCNV
tmp <- subset(iPsychCNV.Eval, CN != 2)
tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
tmp6 <- tapply(tmp$True.Positive, as.factor(tmp$Heterozygosity), function(X){ sum(X)/length(X) })
#df2 <- data.frame(AUC=iPsychCNV.AUC,True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="iPsychCNV", Loop=Loops, stringsAsFactors=F)
df2 <- data.frame(AUC=iPsychCNV.AUC$auc[1], True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5), as.numeric(tmp6)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5)), rep("Heterozygosity", length(tmp6))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5), names(tmp6)), Source="iPsychCNV", Loop=Loops, stringsAsFactors=F)
## iPsychCNV + ReScan
#tmp <- subset(ReiPsych.Eval, CN != 2)
#tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
#tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
#tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
#tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
#tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
#tmp6 <- tapply(tmp$True.Positive, as.factor(tmp$Heterozygosity), function(X){ sum(X)/length(X) })
#df3 <- data.frame(AUC=ReiPsych.AUC,True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="iPsychCNV + ReScan", Loop=Loops, stringsAsFactors=F)
#df3 <- data.frame(AUC=ReiPsych.AUC, True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5), as.numeric(tmp6)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5)), rep("Heterozygosity", length(tmp6))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5), names(tmp6)), Source="iPsychCNV + ReScan", Loop=Loops, stringsAsFactors=F)
## PennCNV + Filter
#tmp <- subset(Filter.Eval, CN != 2)
#tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
#tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
#tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
#tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
#tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
#df3 <- data.frame(AUC=Filter.AUC, True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="Filter", Loop=Loops, stringsAsFactors=F)
## PennCNV + Filter + Hotspots + ReScan
#tmp <- subset(Rescan.Eval, CN != 2)
#tmp$True.Positive <- tmp$CNV.Present == tmp$CNV.Predicted
#tmp2 <- tapply(tmp$True.Positive, as.factor(tmp$sd), function(X){ sum(X)/length(X) })
#tmp3 <- tapply(tmp$True.Positive, as.factor(tmp$CNVmean), function(X){ sum(X)/length(X) })
#tmp4 <- tapply(tmp$True.Positive, as.factor(tmp$NumSNPs), function(X){ sum(X)/length(X) })
#tmp5 <- tapply(tmp$True.Positive, as.factor(tmp$CN), function(X){ sum(X)/length(X) })
#df4 <- data.frame(AUC=Rescan.AUC, True.positive=c(as.numeric(tmp2), as.numeric(tmp3), as.numeric(tmp4), as.numeric(tmp5)), By=c(rep("sd", length(tmp2)), rep("CNV.mean", length(tmp3)), rep("NumSNPs", length(tmp4)), rep("CN", length(tmp5))), Values=c(names(tmp2), names(tmp3), names(tmp4), names(tmp5)), Source="ReScan", Loop=Loops, stringsAsFactors=F)
df5 <- rbind(df, df2)
#df5 <- rbind(df, df2, df3)
#df5 <- rbind(df, df2, df3, df4)
system("rm -f MockSample_*")
return(ValuesDF)
})
Res <- MatrixOrList2df(Res)
#Res$Values <- as.numeric(Res$Values)
return(Res)
}
# Res <- MultipleMockData(NSamples=30, NLoops=30)
# ggplot(Res, aes(x=as.factor(Values), y=True.positive)) + geom_boxplot(aes(fill=as.factor(Source)), outlier.shape=NA) + facet_wrap(~ By, scales = "free") + scale_fill_brewer(palette="Set1")
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.