EXPRESSO: Drug Response Prediction Framework

Data availability: The data/ directory is not included in this repository. It is archived separately on Zenodo (DOI: 10.5281/zenodo.21878615) and must be downloaded and unpacked into your local clone — see Downloading the Data below.

Overview

EXPRESSO (EXpression-Profile-RESponSe-Optimizer) is a drug-specific supervised LASSO logistic regression framework for transcriptome-based treatment response prediction. It evaluates predictive power of drug targets and biomarkers across patient cohorts using a Leave-One-Cohort-Out Cross-Validation (LOCOCV) strategy.

Two model variants are implemented:

EXPRESSO-T: LASSO model with the known drug target gene(s) as penalty-free (regularization-exempt) features, with the rest of the transcriptome included as penalized predictors.
EXPRESSO-B: Extends EXPRESSO-T by additionally including context-specific biomarkers — identified via a nested LOCOCV differential expression procedure — as penalty-free features alongside the target gene(s).

Two baseline models are included for comparison:

Vanilla LASSO (noTarget mode): Supervised LASSO with no biological prior; the full transcriptome is penalized equally.
Unsupervised target-only: Ranks patients by the expression of the target gene alone, with no model training.

Repository Structure

project/
│
├── expresso_all.R                         # Main LOCOCV script (EXPRESSO-T / -B / vanilla)
├── expresso_biomarker_search.R            # Nested LOCOCV biomarker discovery (EXPRESSO-B)
├── gene_signature_benchmarking.R          # Gene signature benchmarking (LOCOCV)
├── expressoT_MLmodels_generate.py         # Train/evaluate alternative ML models (EXPRESSO-T context)
├── expressoB_MLmodels_generate.py         # Train/evaluate alternative ML models (EXPRESSO-B context)
├── Methods_benchmarking.ipynb             # Jupyter notebook: Other methods benchmarking walkthrough
├── source_EXPRESSO_func.R                 # All shared R helper functions
├── renv.lock                              # Pinned R package versions (see Software Requirements)
├── renv/                                  # renv scaffolding (activate.R, settings.json)
├── SESSION_INFO.txt                       # sessionInfo() from the original analysis environment
├── environment.yml                        # Conda environment for the Python benchmarking scripts
├── spec-file.txt                          # Exact conda package spec (alternate reproducible install)
│
├── prospective/                           # Prospective / independent validation pipeline
│   ├── source_EXPRESSO_func.R             # All shared R helper functions
│   ├── expresso_prospective_predict.R     # Get prediction for new cohort(s)
│   ├── expresso_single_cohort_predict.sh  # Shell script to run full prospective pipeline
│   ├── pooled_models/                     # Pooled models trained on all original cohorts, per drug
│   ├── prospec_data/                      # Independent/prospective test cohorts -NOT included in this repo — download from Zenodo (DOI: 10.5281/zenodo.21878615 (https://doi.org/10.5281/zenodo.21878615)) (same layout as data/)
│   └── expresso_test_prospec/             # Prediction outputs (AUC summaries) written here
│
├── data/                                   # NOT included in this repo — download from Zenodo
│   │                                       # (see "Downloading the Data" below); layout once unpacked:
│   ├── <cohort>/                          # One directory per drug cohort group
│   │   ├── metadata.tsv                   # Cohort metadata (cohort, cancer, type, drug, targets)
│   │   │                                  # TRAINING COHORTS ONLY — test cohorts must be excluded
│   │   ├── response.tsv                   # Sample-level treatment response labels
│   │   └── mrna/
│   │       └── <cohort_id>.rds            # Expression matrix (genes × samples, numeric matrix)
│   ├── biomarkers/
│   │   └── <cohort>_biomarkers.tsv        # Per-fold target/biomarker gene definitions
│   └── MultiAssayExperiment/
│       └── MAE_<drug>.rds                 # Publicly available cohorts as MAE objects, per drug
│
├── comparisons/
│   ├── final_signatures.tsv               # Index of all 16 published gene signatures
│   ├── <signature_name>.tsv               # One file per gene signature (gene lists)
│   └── ...
│
└── results/                               # All output TSV files written here

prospective/pooled_models/ is included directly in this repository. The top-level data/ directory (original training cohorts, biomarker gene lists, and MultiAssayExperiment objects) and prospective/prospec_data/ are hosted separately on Zenodo due to its size — see Downloading the Data below.

Downloading the Data

The data/ directory — cohort-level expression matrices, response labels, biomarker gene lists, and MultiAssayExperiment objects — is archived separately on Zenodo rather than committed to this repository.

Zenodo record (DOI): 10.5281/zenodo.21878615
Archive file: expresso_aug2026.tar.gz

prospective/pooled_models/ and prospective/prospec_data/ are already included in this GitHub repository and do not require this step — only the top-level data/ directory needs to be downloaded.

Steps

Clone this repository, if you haven't already, and cd into it:

git clone https://github.com/ruppinlab/EXPRESSO.git
cd EXPRESSO

Download the archive from Zenodo:
```
curl -L -o expresso_aug2026.tar.gz \
  "https://zenodo.org/records/21878615/files/expresso_aug2026.tar.gz?download=1"
```
(or use wget -O expresso_aug2026.tar.gz "<same URL>", or download it directly from the Zenodo record page in a browser)
Extract the archive:
```
tar -xzf expresso_aug2026.tar.gz
```
Move the extracted data/ directory into the repository root, so it sits alongside comparisons/ and prospective/. Place the prospective/prospec_data/ directory for validation cohorts. Check where extraction placed it first — some archive tools unpack into their own top-level folder (e.g. expresso_aug2026/data/) rather than directly into the current directory:
```
ls                                   # confirm where data/ landed after extraction
mv expresso_aug2026/data ./data      # adjust the source path if it differs
```

Confirm the directory landed in the right place:

ls data
# beva  biomarkers  brafi  cyclop  MultiAssayExperiment  pacli  pd1mel  pd1oth  tras

(Optional) remove the downloaded archive once extraction is confirmed:
```
rm expresso_aug2026.tar.gz
```

Once data/ is in place, the repository matches the structure described throughout this README, and the pipeline can be run as documented below.

Input Data Format

`metadata.tsv`

Column	Description
`cohort`	Unique cohort identifier (e.g. `GSE78220`)
`cancer`	Cancer type
`type`	Sample type
`drug`	Drug/intervention name
`targets`	Drug target gene(s)

`response.tsv`

Column	Description
`cohort`	Cohort identifier (must match `metadata.tsv`)
`sample`	Sample ID (must match column names in expression matrix)
`response`	`Responder` or `Non-responder`

`mrna/<cohort_id>.rds`

A numeric R matrix with genes as row names and sample IDs as column names. Gene expression values should be in raw or log-scale; normalization is applied internally (see Normalization section).

`data/biomarkers/<cohort>_biomarkers.tsv`

Column	Description
`test_id`	Cohort held out as test set for this fold
`target`	Comma-separated target gene(s) for `target` mode
`biomarkers`	Comma-separated biomarker genes for `biomarker` mode

Normalization

Raw gene expression values are rank-normalized within each sample (converting expression values to fractional ranks in [0, 1]) using the rank_normalization() function in source_EXPRESSO_func.R. This within-sample normalization is applied uniformly across all cohorts prior to model training.

The --normalization argument (default: "ranked") controls this behavior:

Value	Description
`ranked`	Within-sample rank normalization to [0,1] — default, recommended
`NPN`	Nonparanormal transformation (rank + probit)
`raw`	No normalization applied

Software Requirements

R (≥ 4.0)

This repository includes an renv.lock file (plus renv/activate.R and renv/settings.json) that pins the exact R package versions used for the analysis. Recommended setup:

install.packages("renv")
renv::restore()   # reads renv.lock and installs the matching package versions

SESSION_INFO.txt records the full sessionInfo() output from the original analysis environment as an additional reference if you need to check a specific package version.

Alternatively, install the required packages manually:

install.packages(c(
  "tidyverse", "data.table", "glmnet", "caret",
  "rsample", "pROC", "limma", "parallel", "metap"
))

# For differential expression:
if (!requireNamespace("BiocManager", quietly = TRUE))
  install.packages("BiocManager")
BiocManager::install("limma")

Python (3.11)

pip install numpy pandas scipy matplotlib

This repository also includes environment.yml and spec-file.txt for reproducing the exact conda environment used for the benchmarking scripts:

conda env create -f environment.yml
# or, for an exact package-for-package match:
conda create --name expresso --file spec-file.txt

Random Seeds

All stochastic procedures use fixed random seeds for full reproducibility:

set.seed(31052024) is called immediately before each cv.glmnet() model build in source_EXPRESSO_func.R
The bootstrap CI functions in the Python comparison scripts use np.random.default_rng(42)

Scripts and Usage

1. `expresso_all.R` — Main LOCOCV (EXPRESSO-T / -B / noTarget)

Runs LOCOCV for a single drug and produces AUCs, odds ratios, and selected features.

Rscript expresso_all.R <intervention> <cohort> <biomarker_file> <mode> [normalization] [response_weights]

Argument	Description
`intervention`	Drug name (e.g. `ICB`, `trastuzumab`)
`cohort`	Cohort directory under `data/` (e.g. `pd1mel`)
`biomarker_file`	TSV file under `data/biomarkers/` with per-fold target/biomarker genes
`mode`	`target` (EXPRESSO-T), `biomarker` (EXPRESSO-B), or `noTarget`
`normalization`	(optional) `ranked` (default), `NPN`, or `raw`
`response_weights`	(optional) `TRUE` (default) or `FALSE`

Drug-cohort reference table:

Drug	intervention	cohort	biomarker_file
ICB – melanoma	`ICB`	`pd1mel`	`pd1mel_biomarkers.tsv`
ICB – non-melanoma	`ICB`	`pd1oth`	`pd1oth_biomarkers.tsv`
Trastuzumab	`trastuzumab`	`tras`	`tras_biomarkers.tsv`
Bevacizumab	`bevacizumab`	`beva`	`beva_biomarkers.tsv`
BRAFi	`BRAFi`	`brafi`	`brafi_biomarkers.tsv`
Paclitaxel	`paclitaxel`	`pacli`	`pacli_biomarkers.tsv`
Chemo-FAC-FEC	`cyclophos`	`cyclop`	`cyclop_biomarkers.tsv`

Examples:

# EXPRESSO-T for Chemo-FAC-FEC
Rscript expresso_all.R cyclophos cyclop cyclop_biomarkers.tsv target

# EXPRESSO-B for ICB-melanoma
Rscript expresso_all.R ICB pd1mel pd1mel_biomarkers.tsv biomarker

# Vanilla LASSO (no target prior) for paclitaxel
Rscript expresso_all.R paclitaxel pacli pacli_biomarkers.tsv noTarget

Outputs written to results/:

File	Contents
`<intervention>_<cohort>_<mode>_expresso_AUCs.tsv`	Per-cohort AUC and odds ratio (mid.OR), plus Mean/Median summary rows
`<intervention>_<cohort>_<mode>_expresso_features.tsv`	Genes selected per fold with LASSO coefficients
`<intervention>_<cohort>_<mode>_expresso_features_frequency.tsv`	Aggregated gene selection frequency across all folds

Note on filename convention for downstream comparison scripts: The Python comparison scripts (expressoT_MLmodels_generate.py, expressoB_MLmodels_generate.py) expect AUC files named <intervention>_LOCOCV_lasso_AUCs.tsv. Rename or symlink the output from expresso_all.R accordingly before running comparison scripts.

2. `expresso_biomarker_search.R` — Biomarker Discovery (Nested LOCOCV)

Runs the nested LOCOCV biomarker discovery procedure used to produce the per-fold biomarker gene lists for EXPRESSO-B. For each outer fold, identifies differentially expressed genes from the training cohorts using limma, combines fold-level p-values using Brown's method (to account for inter-fold dependence from overlapping training sets), and selects genes that improve LOCOCV AUC by ≥ 0.02 (ΔAUC criterion).

Rscript expresso_biomarker_search.R <intervention> <cohort> <target> [options]

Argument	Description
`intervention`	Drug name
`cohort`	Cohort identifier
`target`	Target gene(s), comma-separated (e.g. `CD274` or `TOP2A,TYMS`)

Key options:

Option	Default	Description
`--de_fdr`	`0.05`	FDR cutoff for DE gene selection
`--de_logfc`	`0.4`	Minimum mean absolute log-fold-change
`--delta_auc`	`0.02`	Minimum ΔAUC to accept a biomarker gene
`--delta_pval`	`0.05`	Wilcoxon p-value threshold for ΔAUC test
`--max_biom`	`3`	Maximum biomarker genes to add per fold
`--ncores`	auto	Parallel cores for deltaAUC computation
`--result_dir`	`<cohort>_cv_fold_results`	Output directory

Examples:

Rscript expresso_biomarker_search.R ICB pd1mel CD274
Rscript expresso_biomarker_search.R trastuzumab tras ERBB2 --ncores 8 --max_biom 3

Outputs written to --result_dir:

File	Contents
`<intervention>_<cohort>_cv_result_<test_id>.rds`	Per-fold RDS with DE summary, deltaAUC results, selected biomarker genes
`<intervention>_<cohort>_summary_auc_results.tsv`	Summary AUC table across all folds

The per-fold RDS files are used to populate the biomarkers column in <cohort>_biomarkers.tsv for input to expresso_all.R.

3. `gene_signature_benchmarking.R` — Published Signature Benchmarking

Evaluates a set of published gene signatures against all cohorts in a prospective (non-cross-validated) manner. Each signature is scored by the mean rank-normalized expression of its member genes; AUC is computed per cohort.

Rscript gene_signature_benchmarking.R <intervention> <cohort> <signatures_file> [options]

Argument	Description
`intervention`	Drug name (used in output filename)
`cohort`	Cohort identifier
`signatures_file`	TSV file with a `signatures` column listing signature names

Key options:

Option	Default	Description
`--sig_dir`	`comparisons`	Directory containing per-signature TSV gene list files
`--result_dir`	`comparisons`	Directory for output TSV
`--drug_label`	same as `intervention`	Human-readable label in output table

Example:

Rscript gene_signature_benchmarking.R trastuzumab tras comparisons/final_signatures.tsv \
  --sig_dir comparisons --result_dir results --drug_label "Trastuzumab"

Output: <intervention>_<cohort>_aucs_comparison_gene_signatures_long_loco_pros.tsv

Contains columns: cohort, AUC, drug, predictor (signature name).

The comparisons/ directory contains gene lists for all 16 published signatures used for benchmarking (e.g., TIDE, IMPRES, CYT, MammaPrint, OncotypeDX), each as a separate TSV file, along with references to the original publications. Signatures were implemented according to their original scoring procedures using the same rank-normalized expression data as EXPRESSO.

4. `expressoT_MLmodels_generate.py` / `expressoB_MLmodels_generate.py` — Alternative ML Models

Generate LOCOCV AUC results for alternative machine learning models (Random Forest, XGBoost, SVM, KNN, MLP) in the EXPRESSO-T and EXPRESSO-B contexts respectively, for direct comparison with EXPRESSO. Both scripts also run paired Wilcoxon signed-rank tests against the LASSO reference, compute bootstrap 95% confidence intervals, and produce a publication-ready boxplot.

Output files follow the naming convention:

results/<intervention>_LOCOCV_<model>_AUCs.tsv

python expressoT_MLmodels_generate.py \
  --results_dir results \
  --intervention ICBmel \
  --reference lasso \
  --output_dir figures

python expressoB_MLmodels_generate.py \
  --results_dir results \
  --intervention ICBmel \
  --reference lasso \
  --output_dir figures

Option	Default	Description
`--results_dir`	(required)	Directory with `<intervention>_LOCOCV_<model>_AUCs.tsv` files
`--intervention`	(required)	Drug name matching the file prefix
`--reference`	`lasso`	Reference model for Wilcoxon test
`--min_cohort_size`	`0`	Minimum cohort size filter
`--output_dir`	`figures`	Directory for output PDF/PNG and stats TSV
`--n_boot`	`2000`	Bootstrap iterations for CI computation

Outputs:

File	Contents
`<intervention>_model_stats.tsv`	Mean AUC, 95% CI, Wilcoxon statistic, p-value, significance per model
`<intervention>_boxplot_AUC.pdf/.png`	Boxplot with per-cohort AUCs, jittered points, and significance stars

5. `Methods_benchmarking.ipynb` — Walkthrough Notebook

A Jupyter notebook demonstrating a full end-to-end run of EXPRESSO-T and EXPRESSO-B for a single drug and cohort, including data loading, model training, AUC evaluation, and signature benchmarking. Intended as a reproducibility reference and example for new users.

Prospective Validation

The prospective/ directory contains scripts for applying a trained EXPRESSO model to new, completely unseen cohorts. Pre-trained pooled models (pooled_models/) and the independent/prospective cohort data (prospec_data/) are already included under prospective/, so reproducing the manuscript's prospective results only requires the prediction step below — no separate model training or download is needed.

Data organisation for prospective validation

data/
└── <cohort>/
    ├── metadata.tsv    ← TRAINING cohorts only (test cohorts must be excluded)
    ├── response.tsv
    └── mrna/

prospective/
├── prospec_data/
│   └── <cohort>/
│       ├── metadata.tsv    ← TEST cohorts only
│       ├── response.tsv
│       └── mrna/
│           └── <cohort_id>.rds
│
├── pooled_models/          ← pooled model trained on all original cohorts, per drug
│   ├── <intervention>_<cohort>_target_expresso_pooled_model.rds
│   ├── <intervention>_<cohort>_target_gene_list.tsv
│   └── <intervention>_<cohort>_target_scale_params.rds   ← training-data scaling params, applied to new cohorts before prediction
│
└── expresso_test_prospec/  ← prediction outputs written here

Important: data/<cohort>/metadata.tsv must contain training cohorts only. Test cohorts must be completely absent from this file. Mixing them in will produce a model that has seen the test data during training, invalidating the prospective evaluation.

6. `prospective/expresso_prospective_predict.R` — Predict on New Cohort(s)

Loads a saved pooled model and evaluates it on one or more new cohorts. Uses summarize_ORs() from source_EXPRESSO_func.R exactly as the training pipeline does, so AUC and OR values are directly comparable to LOCOCV results.

Rscript prospective/expresso_prospective_predict.R \
  <model_rds> <gene_list_tsv> <data_dir> [cohort_names] [normalization]

Argument	Description
`model_rds`	Path to pooled model RDS, e.g. `prospective/pooled_models/<...>_expresso_pooled_model.rds`
`gene_list_tsv`	Path to the matching gene list TSV in the same directory
`data_dir`	Directory with `metadata.tsv`, `response.tsv`, `mrna/` for test cohorts
`cohort_names`	(optional) comma-separated cohort IDs to evaluate — default: all in directory
`normalization`	(optional) `ranked` (default) — must match what was used during training

The matching <...>_scale_params.rds file in pooled_models/ is loaded automatically alongside the model to standardize new-cohort expression consistently with training.

Examples:

# Evaluate all cohorts in the prospective directory
Rscript prospective/expresso_prospective_predict.R \
  prospective/pooled_models/trastuzumab_tras_target_expresso_pooled_model.rds \
  prospective/pooled_models/trastuzumab_tras_target_gene_list.tsv \
  prospective/prospec_data/tras

# Evaluate specific cohorts only
Rscript prospective/expresso_prospective_predict.R \
  prospective/pooled_models/ICB_pd1oth_target_expresso_pooled_model.rds \
  prospective/pooled_models/ICB_pd1oth_target_gene_list.tsv \
  prospective/prospec_data/pd1oth \
  GSE281729_p1,GSE281729_p2,GSE274975

Output written to prospective/expresso_test_prospec/:

File	Contents
`<intervention>_<cohort>_<mode>_<dir>_AUC_summary.tsv`	Per-cohort AUC and OR (mid.OR), plus Mean/Median summary rows

6b. `prospective/expresso_single_cohort_predict.sh` — Run Full Prospective Pipeline

Shell wrapper that runs expresso_prospective_predict.R for all drug/cohort combinations below in one call:

bash prospective/expresso_single_cohort_predict.sh

Drug-cohort reference for prospective validation:

Drug	intervention	cohort	Test cohorts
ICB – melanoma	`ICB`	`pd1mel`	`davar24`
ICB – non-melanoma	`ICB`	`pd1oth`	`GSE281729_p1`, `GSE281729_p2`, `GSE274975`
ICB – lung (NSCLC-specific sub-model of `pd1oth`)	`ICB`	`icblung`	`GSE274975`
Trastuzumab	`trastuzumab`	`tras`	`GSE130788`, `GSE243375`, `phs003576`
Chemo-FAC-FEC	`cyclophos`	`cyclop`	`GSE122630`, `GSE123845_p2`, `GSE14764`, `GSE16716`, `GSE21974`, `GSE22226`, `GSE231629`, `GSE240671_p2`, `GSE25066`, `GSE260693`, `GSE32603`, `GSE34138`, `GSE41656`, `GSE42822_p2`, `GSE4779`

Typical Full Workflow

Step 1: Biomarker discovery (produces per-fold biomarker gene lists)
  └─ expresso_biomarker_search.R  →  <cohort>_biomarkers.tsv

Step 2: LOCOCV model evaluation
  ├─ expresso_all.R (mode=target)    →  EXPRESSO-T AUCs + features
  ├─ expresso_all.R (mode=biomarker) →  EXPRESSO-B AUCs + features
  └─ expresso_all.R (mode=noTarget)  →  Vanilla LASSO AUCs

Step 3: Alternative ML model evaluation
  ├─ expressoT_MLmodels_generate.py  →  RF, XGB, SVM, KNN, MLP AUCs (T context)
  └─ expressoB_MLmodels_generate.py  →  RF, XGB, SVM, KNN, MLP AUCs (B context)

Step 4: Signature benchmarking
  ├─ gene_signature_benchmarking.R   →  Published signature AUCs per cohort
  └─ Methods_benchmarking.ipynb      →  Interactive benchmarking walkthrough

Step 5: Statistical comparison and visualization
  ├─ expressoT_MLmodels_generate.py  →  EXPRESSO-T boxplot vs ML models
  └─ expressoB_MLmodels_generate.py  →  EXPRESSO-B boxplot vs ML models

Step 6: Prospective validation (pooled models & test cohorts already included)
  ├─ prospective/expresso_prospective_predict.R      →  Predict on new cohort(s)
  └─ prospective/expresso_single_cohort_predict.sh   →  Run all drug/cohort combinations at once

Notes

The helper functions in source_EXPRESSO_func.R must be in the same directory as the calling script, or the --source_file path must be updated.
Cohorts containing dbGaP-restricted data require data access approval; results for those cohorts will differ from the manuscript until access is granted.
Genes are restricted to those measured in at least 3 cohorts (controlled by gene_cohort_cutoff_) to enable consistent cross-cohort model fitting.
For prospective validation, data/<cohort>/metadata.tsv must contain training cohorts only. Test cohorts live in prospective/prospec_data/<cohort>/metadata.tsv.
The gene list embedded in the saved model (read from rownames(coef(mod))) is the authoritative source of gene ordering for prediction — it takes precedence over the gene_list.tsv file.
The full list of cohorts used per drug is provided in Supplementary Table S1 of the manuscript.

Navigation Menu

Uh oh!

Repository files navigation

EXPRESSO: Drug Response Prediction Framework

Overview

Repository Structure

Downloading the Data

Steps

Input Data Format

`metadata.tsv`

`response.tsv`

`mrna/<cohort_id>.rds`

`data/biomarkers/<cohort>_biomarkers.tsv`

Normalization

Software Requirements

R (≥ 4.0)

Python (3.11)

Random Seeds

Scripts and Usage

1. `expresso_all.R` — Main LOCOCV (EXPRESSO-T / -B / noTarget)

2. `expresso_biomarker_search.R` — Biomarker Discovery (Nested LOCOCV)

3. `gene_signature_benchmarking.R` — Published Signature Benchmarking

4. `expressoT_MLmodels_generate.py` / `expressoB_MLmodels_generate.py` — Alternative ML Models

5. `Methods_benchmarking.ipynb` — Walkthrough Notebook

Prospective Validation

Data organisation for prospective validation

6. `prospective/expresso_prospective_predict.R` — Predict on New Cohort(s)

6b. `prospective/expresso_single_cohort_predict.sh` — Run Full Prospective Pipeline

Typical Full Workflow

Notes

About

Releases

Packages

Contributors

Languages

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
comparisons		comparisons
prospective		prospective
renv		renv
Methods_benchmarking.ipynb		Methods_benchmarking.ipynb
README.md		README.md
README.txt		README.txt
SESSION_INFO.txt		SESSION_INFO.txt
environment.yml		environment.yml
expressoB_MLmodels_generate.py		expressoB_MLmodels_generate.py
expressoT_MLmodels_generate.py		expressoT_MLmodels_generate.py
expresso_all.R		expresso_all.R
expresso_biomarker_search.R		expresso_biomarker_search.R
gene_signature_benchmarking.R		gene_signature_benchmarking.R
renv.lock		renv.lock
source_EXPRESSO_func.R		source_EXPRESSO_func.R
spec-file.txt		spec-file.txt

Uh oh!

Folders and files

Latest commit

History

Repository files navigation

EXPRESSO: Drug Response Prediction Framework

Overview

Repository Structure

Downloading the Data

Steps

Input Data Format

metadata.tsv

response.tsv

mrna/<cohort_id>.rds

data/biomarkers/<cohort>_biomarkers.tsv

Normalization

Software Requirements

R (≥ 4.0)

Python (3.11)

Random Seeds

Scripts and Usage

1. expresso_all.R — Main LOCOCV (EXPRESSO-T / -B / noTarget)

2. expresso_biomarker_search.R — Biomarker Discovery (Nested LOCOCV)

3. gene_signature_benchmarking.R — Published Signature Benchmarking

4. expressoT_MLmodels_generate.py / expressoB_MLmodels_generate.py — Alternative ML Models

5. Methods_benchmarking.ipynb — Walkthrough Notebook

Prospective Validation

Data organisation for prospective validation

6. prospective/expresso_prospective_predict.R — Predict on New Cohort(s)

6b. prospective/expresso_single_cohort_predict.sh — Run Full Prospective Pipeline

Typical Full Workflow

Notes

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

`metadata.tsv`

`response.tsv`

`mrna/<cohort_id>.rds`

`data/biomarkers/<cohort>_biomarkers.tsv`

1. `expresso_all.R` — Main LOCOCV (EXPRESSO-T / -B / noTarget)

2. `expresso_biomarker_search.R` — Biomarker Discovery (Nested LOCOCV)

3. `gene_signature_benchmarking.R` — Published Signature Benchmarking

4. `expressoT_MLmodels_generate.py` / `expressoB_MLmodels_generate.py` — Alternative ML Models

5. `Methods_benchmarking.ipynb` — Walkthrough Notebook

6. `prospective/expresso_prospective_predict.R` — Predict on New Cohort(s)

6b. `prospective/expresso_single_cohort_predict.sh` — Run Full Prospective Pipeline