Using R to interface with the cBioPortal

README.md

#Using R to interface with the cBioPortal#

cgdsr.r

library (cgdsr)

# Create a connection to the cBioPortal for Cancer Genomics
mycgds = CGDS("http://www.cbioportal.org/public-portal/")

# Test the connection
test(mycgds)

# Instead of going through the 80-something studies,
# let's focus on the TCGA datasets (shown in the text file below)
library(RCurl)
url = getURL("http://bit.ly/tcga_study_ids", followlocation = TRUE)
tcga_studies = as.matrix(read.table(text = url))

# Create a list of cancer-related genes
# The names of the genes must follow HUGO Gene Nomenclature Committee
# HGNC (http://www.genenames.org/)
genes = c("EGFR", "HER1", "MYC", "RB1", "TP53", "PTEN")

# Empty dataset for holding the expression values
exps = NULL

# For each study:
for (i in 1 : length(tcga_studies)) {
	# Obtain the study id
	study_id = tcga_studies[i]
	
	# Build the mRNA profile (based on the study id)
	profile_mrna = paste(study_id, "rna_seq_v2_mrna_median_Zscores", sep = "_")
	
	# Select the "completely" analyzed cancer samples
	# i.e., mRNA expression, copy number variation, and mutations
	case = paste(study_id, "3way_complete", sep = "_")
	
	# Empty vector for the means of the genes in the current study
	means = NULL
	
	# For each gene:
	for (j in 1 : length(genes)) {
		# Retrieve expression data from the server
		mrna = getProfileData(mycgds, genes[j], profile_mrna, case)
		
		# Compute the mean of the expression values for the current gene
		means = c(means, mean(mrna[,1]))
	}

	# Add a row to the table of the genes
	exps = rbind(exps, means)
}

row.names(exps) = tcga_studies
colnames(exps) = genes

# Boxplot for the means of the genes of interest in the different cancer types
boxplot (exps)

tcga_studies.txt

acc_tcga
blca_tcga
blca_tcga_pub
brca_tcga
cesc_tcga
coadread_tcga
gbm_tcga
gbm_tcga_pub2013
hnsc_tcga
hnsc_tcga_pub
kich_tcga
kirc_tcga
kirc_tcga_pub
kirp_tcga
laml_tcga
laml_tcga_pub
lgg_tcga
luad_tcga
luad_tcga_pub
lusc_tcga
ov_tcga
paad_tcga
prad_tcga
skcm_tcga
stad_tcga
stad_tcga_pub
thca_tcga
ucec_tcga_pub