Information about the repository

This directory contains all computer code to perform the analyses from our upcoming manuscript, "A polygenic score-based approach to identify gene-drug interactions stratifying breast cancer risk".

June 14, 2021: Please note that the scripts could use further commenting and documentation. This is on the to-do list. Currently, this mostly serves as a repository for myself to access any relevant scripts but other researchers can access these scripts in their current format as well.

All scripts are written in the R or Bash programming languages. The analysis was performed on a linux system. Plots were created on a macOS Catalina system.

If you can not find the code you are looking for or have any questions, please contact:

Andrew Marderstein [email protected]

If you use our work, please cite:

Marderstein, A.R., Kulm, S., Peng, C., Tamimi, R.M., Clark, A.G.^, Elemento, O.^ (2021). A polygenic score-based approach to identify gene-drug interactions stratifying breast cancer risk. medRxiv. doi: https://doi.org/10.1101/2021.05.03.21256511.

Generating the analysis data

The create_data directory contains data-generating files.

Calculating polygenic scores

Please see the create_score and create_score_PGS_catalog directories.

Polygenic score metrics

pgs_select_optimal_score_plot_plus_heatmap.R
pgs_corr_plot.R
identify_best_score.R
cv_overfit.R
interaction_analysis_plots.R

Polygenic-drug interaction tests

pgs_drug_gxdrug.R
pgs_qq.R
variation_explained_by_gxdrug.R
pgs_variation_explained.R

Single-SNP interaction analyses

Creating genotype files for the single-SNP analysis using:

create_genotype_files_for_snp_gxe.sh
snp_drug_gxdrug.R
snp_drug_gxdrug.analyze_res.R
coord_int_statistics.R
create_snpxcort_supp_tab.R
survival_plots.R

In silico single-SNP functional analyses:

ukb-id_to_open-targets-id.sh
enrichment_analysis.R
enrichr_analyses.R
gtex_snp.R

Sensitivity analyses for confounders:

confounding_analyses.R
corticosteroid_indic_create.R
sensitivity_remove-indic.R

Other scripts

figure1.R