OnCorr: A pan-cancer mRNA-protein correlation tool for precision oncology

Overview

OnCorr is an interactive web tool for exploring the relationship between mRNA and protein abundance across multiple large-scale pan-cancer datasets. While transcriptomics is often used as a proxy for protein levels, mRNA–protein correlations vary widely across genes, tissues, and biological contexts. OnCorr makes this variability explicit by enabling users to systematically examine gene-specific correlations across multiple cohorts and tissue types.

The repository contains the code used for data processing, correlation analysis, and figure generation for the OnCorr manuscript.

Repository structure

├── analysis
├── code
├── data
│   ├── Datasets
│   │   └── CCLE
│   ├── OncoKb
│   ├── Pathways
│   └── Ranks
├── docs
│   ├── figure
│   │   ├── publication-figures.Rmd
│   │   └── supplementary-publication-figures.Rmd
│   └── site_libs
└── output
    ├── correlations-datasets
    └── Figures

analysis/:

Contains analysis scripts and notebooks used to explore mRNA–protein correlations, perform downstream statistical analyses, and generate intermediate results used in the manuscript.

code/:

Core functions and reusable code for data processing, correlation calculations, filtering, and annotation. These scripts underpin all analyses and figure generation in the repository.

data/:

Input data and reference resources used throughout the project, including processed datasets, pathway annotations, external knowledge bases, and ranked gene-level results required for analysis.

docs/:

R Markdown into html files and supporting assets used to generate publication and supplementary figures. This directory contains the source files for all figures reported in the manuscript.

output/:

Generated outputs from analyses, including correlation results and final figures. All files in this directory are derived from scripts in analysis/ and code/.

How to regenerate figures or analysis

All analyses and figures in the OnCorr manuscript can be reproduced using the code in this repository.

Requirements Analyses were performed in R (≥ 4.2). Required R packages are listed within the analysis and figure-generation scripts. For full reproducibility, we recommend running the code in a clean R environment.

Data Input datasets and reference resources are provided in the data/ directory. These include pathway annotations from , external knowledge bases, and precomputed gene rankings from used throughout the analyses.

The processed transcriptomic and proteomic datasets can be retrieved from:

Dataset	Proteomic location	Transcriptomic location
CCLE	Nusinow et al	Nusinow et al
CPTAC	LinkedOmicsKB	LinkedOmicsKB
ProCan-DepMapSanger	Gonçalves et al	Cell Model Passports

Analysis: All statistical analyses and correlation calculations are implemented in the analysis/ directory. Running the scripts in analysis/calculate-correlations.Rmd will regenerate the correlation results written to output/correlations-datasets. Please ensure that the paths to input data are correctly specified before running the analysis.

Figures All manuscript and supplementary figures are generated from R Markdown documents located in analysis/publication-figures.Rmd and analysis/supplementary-figures.Rmd. Rendering these files will reproduce the figures saved to output/Figures.