PharmacoGx

PharmacoGx is an R/Bioconductor package we developed to simplify the analysis of large PharmacoGenomic Datasets. All the datasets included in PharmacoDB are available as R objects for analysis using PharmacoGx, with both the pharmacological and molecular data from each study. PharmacoGx incorporates all the annotations of each study available in PharmacoDB, and provides a suite of statistical modeling functions to jointly analyze molecular features and drug dose-response curves.

PharmacoGx is simple to start using once R is installed and running on your system. If its your first time using PharmacoGx, download and install the package and all its dependencies by opening an R console and running the following commands:

source("http://www.bioconductor.org/biocLite.R")
biocLite("PharmacoGx")

After installing for the first time, to load the package you can run the following command in your R session:

library(PharmacoGx)

To download a PharmacoSet object with all the pharmacological and molecular data for UHNBreast and begin analysing the data, you can use the following commands.

# Download the object for UHNBreast
UHNBreast <- downloadPSet(UHNBreast)

# Plot Drug Dose response curves, using the same names for compounds and cell lines as PharmacoDB
drugDoseResponseCurve(UHNBreast, drug="paclitaxel", cell="MCF7")

# Extract the expression data to a matrix
UHNBreast.expression <- summarizeMolecularProfiles(UHNBreast, mDataType="rna")

# Run a linear model for univariate biomarker discovery
UHNBreast.sensitivity.signatures <- drugSensitivitySig(UHNBreast, mDataType="rna", sensitivity.measure="auc_recomputed")

To learn more about the PharmacoGx package, you can download the full documentation and vignettes through our Bioconductor page: PharmacoGx