TargetScore is a data-driven network modeling algorithm for drug response analysis in cancer.
Install Git, R 3.5 or later, and Pandoc. Pandoc is included with RStudio; the Docker validation uses R 4.6.
- Clone the repository and enter its root directory:
git clone https://github.com/korkutlab/targetscore.git
cd targetscore- From the repository root, start R and install the package dependencies and TargetScore:
install.packages(
c("remotes", "rmarkdown"),
repos = "https://cloud.r-project.org"
)
remotes::install_github(
"cmap/morpheus.R",
dependencies = TRUE,
upgrade = "never"
)
remotes::install_deps(
"targetscore",
dependencies = TRUE,
upgrade = "never"
)
remotes::install_local(
"targetscore",
dependencies = FALSE,
upgrade = "never"
)- In the same R session, render the tutorial:
dir.create("output", showWarnings = FALSE)
output_dir <- normalizePath("output")
rmarkdown::render(
"targetscore/vignettes/target_score_tutorial.Rmd",
output_format = "html_document",
output_dir = output_dir,
params = list(output_dir = output_dir)
)- Open
output/target_score_tutorial.htmlin a web browser. All other files created by the tutorial are also retained inoutput/.
Installation requires access to CRAN and the public cmap/morpheus.R GitHub
repository. Some R dependencies may require platform-specific compilers or
system libraries when a prebuilt binary is unavailable.
A normal build runs both the Shiny startup check and the complete vignette pipeline before creating the production image:
docker build --tag targetscore .
docker run --rm --publish 3838:3838 targetscoreOpen http://127.0.0.1:3838/ to use the application. The optional PORT environment variable changes the container's listening port; publish the same port when overriding it.
Build only through the verification stage:
docker build --target verify --tag targetscore-verify .Force Buildx to rerun that stage without invalidating the dependency-installation cache:
docker buildx build --target verify --no-cache-filter verify --load --tag targetscore-verify .Building requires network access to CRAN and the public cmap/morpheus.R GitHub repository. Running the production image requires no credentials, mounted data, external services, or network access.