R Integration

Progenetix API and R

With the ability to access the status matrix directly, one easily can import the data into an R data frame:

pgframe <- read.table(url("http://progenetix.org/api/?icdm_m=814&db=progenetix&api_out=matrix"), header=T, sep="\t", na="NA")

For the segment file, the same applies with “output=segments”:

segtable <- read.table(url("http://progenetix.org/api/?text_m=sezary&db=progenetix&api_out=segments"), header=T, sep="\t", na="NA")

R Example - survival

One can use this example to search for MYCN gain related survival bias in an ICD entity (here “9500/3” - change acc. to your interest). Other modifications are possible.

rm(list = ls())
baseurl  <- "http://progenetix.org/api/?api_out=samples&api_doctype=tab"
db       <- "progenetix"
icdm     <- "9500/3"
gene     <- "MYCN"
qurl     <- paste(baseurl, "&db=", db, "&icdm_m=", icdm, "&genes_m=", gene, "$", sep="")
survData <- read.table(url(qurl), header=T, sep="\t", na="NA")
geneColumn <- paste('GENE_', gene, sep="")
genePos	<- grep(geneColumn, colnames(survData), perl="T")
plot(survfit(Surv(survData$FOLLOWUP, survData$DEATH) ~ 1, se.fit=TRUE), main=paste("Overall survival"), xlab="months", ylab="survival", cex=1.2)
gainNo <- nrow(subset(survData, survData[genePos] == 1))
plot(survfit(Surv(survData$FOLLOWUP, survData$DEATH) ~ survData[genePos] == 1, se.fit=TRUE), col=c("black","blue"), main=paste("Survival and gene ", gene, " (ICD-O ", icdm, ")", sep=""), xlab="months", ylab="survival", cex=1.2)
sdf <- survdiff(Surv(survData$FOLLOWUP, survData$DEATH) ~ survData[genePos] == 1)
pcsq <- round(pchisq(sdf$chisq, df=1, lower=FALSE), digits=5)
legend("bottomright", c("no gain", paste(gene, ' gain (', gainNo, '/', nrow(survData), ')', sep="")), fill=c("black","blue"), inset=c(0.02,0.02), bg="azure1", cex=0.8)
legend("bottomleft", c(paste("p:", pcsq)), inset=c(0.02,0.02), bty="n", cex=1)
Michael Baudis, 2018-05-10

Michael Baudis, 2018-05-10