This Github repo contains the code needed to download and preprocess the data, build the SVM classifier and regenerate all the figures associated with the publication RAS oncogenic activity predicts response to chemotherapy and outcome in lung adenocarcinoma

First step, clone the repo.

git clone https://github.com/FrancisCrickInstitute/RAS84.git
cd RAS84

To run the analysis code you will need R-3.6.0 installed along with the libraries contained in renv.lock. The renv environment can be reinitialised by running R within the RAS84 directory and running renv::restore()

data/scripts contains shell scripts to download each of the dependent datasets. The paths are configured to run these scripts from within data/scripts. Raw data files are downloaded to data/downloads

There are also R scripts to preprocess the raw data and build SummarizedExperiment objects. The SE objects are written to data/objects

The following analysis Rmd scripts can be found in the top level of the repo.

1. 1_build_RAS84_CCLE.Rmd - The CCLE analysis to build RAS84 from the parent signatures

2. 2_drug_screen_GDSC_CTRP_CCLE.Rmd - The CCLE drug screen analysis

3. 3_RAS84_patient_classification_TCGA_LUAD.Rmd - TCGA LUAD classification analysis

4. 4_RAS84_RAG_mutation_analysis_TCGA_LUAD.Rmd - TCGA LUAD RAS Activity Group (RAG) mutation analysis

5. 5_survival_analysis_TCGA_LUAD.Rmd - TCGA LUAD and Uppsala survival analysis

6. 6_RAS84_pancancer_analysis_TCGA.Rmd - TCGA pancancer RAS84 analysis

7. 7_RAS84_pancancer_survival_analysis.Rmd - TCGA pancancer RAS84 survival analysis

8. build_SVM.R - SVM build script

All the Rmd scripts contain global variables in the first code block that set the paths to the required data resources and output paths for the figures. All the defined data resources must be downloaded and build using the download and init scripts described above before the analysis scripts can be run. All figures are written to figures/.