srep-meta-codes

Analysis and visualization code for:

Shakiba M, Rassouli FB. Joining up the scattered anticancer knowledge on auraptene and umbelliprenin: a meta-analysis. Scientific Reports 14, 11770 (2024). https://doi.org/10.1038/s41598-024-62747-z

What this is

Auraptene (AUR) and umbelliprenin (UMB) are prenylated coumarins with reported anticancer activity in cell lines. The paper synthesizes 27 in vitro studies to quantify dose-dependent effects on cell viability, compare the two compounds, and rank which factors (dose, cancer type, treatment duration, compound) most strongly influence toxicity.

This repository contains the Python scripts for the figures and the machine-learning analysis. The extracted study data are archived separately on Dryad.

Data

The dataset is deposited at https://doi.org/10.5061/dryad.cvdncjt9k.

File	Contents
AUR.csv	Extracted observations for auraptene
UMB.csv	Extracted observations for umbelliprenin
combined.csv	AUR and UMB observations combined

Download the files from Dryad and place them in a data/ folder at the repository root:

srep-meta-codes/
├── data/
│   ├── AUR.csv
│   ├── UMB.csv
│   └── combined.csv
└── ...

The column definitions are in the dataset's own README on Dryad.

Repository contents

The analysis scripts are organized by the figure number in the paper:

srep-meta-codes/
├── fig3/
│   ├── AUR-Pie.py
│   └── UMB-Pie.py
├── fig4/
│   ├── AUR-3D-Viab-Dose-Time.py
│   └── UMB-3D-Viab-Dose-Time.py
├── fig5/
│   ├── AUR-Cancer-Viab-Dose.py
│   └── UMB-Cancer-Viab-Dose.py
├── fig6/
│   └── Reg-Scatter-All.py
├── fig7/
│   ├── AUR-Heat-Cor.py
│   └── UMB-Heat-Cor.py
└── fig8/
    ├── RandomForestClassifier.py
    └── RandomForestRegressor.py

Figure	Script	What it produces	Input
3	fig3/AUR-Pie.py	AUR cancer-type distribution pie chart	data/AUR.csv
3	fig3/UMB-Pie.py	UMB cancer-type distribution pie chart	data/UMB.csv
4	fig4/AUR-3D-Viab-Dose-Time.py	AUR 3D scatter plot of time, dose and viability	data/AUR.csv
4	fig4/UMB-3D-Viab-Dose-Time.py	UMB 3D scatter plot of time, dose and viability	data/UMB.csv
5	fig5/AUR-Cancer-Viab-Dose.py	AUR viability versus dose, grouped by cancer type	data/AUR.csv
5	fig5/UMB-Cancer-Viab-Dose.py	UMB viability versus dose, grouped by cancer type	data/UMB.csv
6	fig6/Reg-Scatter-All.py	Viability versus dose scatter plots with regression lines for both compounds and cell lines	data/combined.csv
7	fig7/AUR-Heat-Cor.py	AUR correlation heatmap for dose, treatment time and viability	data/AUR.csv
7	fig7/UMB-Heat-Cor.py	UMB correlation heatmap for dose, treatment time and viability	data/UMB.csv
8	fig8/RandomForestClassifier.py	Random-forest classification metrics and feature-importance plots	data/combined.csv
8	fig8/RandomForestRegressor.py	Random-forest regression metrics and feature-importance plot	data/combined.csv

Installation

git clone https://github.com/moneuron/srep-meta-codes.git
cd srep-meta-codes
python -m venv .venv && source .venv/bin/activate   # optional
pip install -r requirements.txt

Use Python 3 with the packages listed in requirements.txt.

Usage

After placing the data files in data/, run a script from the repository root. Each script displays its plot(s) and does not write an output file:

python fig3/AUR-Pie.py
python fig3/UMB-Pie.py
python fig4/AUR-3D-Viab-Dose-Time.py
python fig4/UMB-3D-Viab-Dose-Time.py
python fig5/AUR-Cancer-Viab-Dose.py
python fig5/UMB-Cancer-Viab-Dose.py
python fig6/Reg-Scatter-All.py
python fig7/AUR-Heat-Cor.py
python fig7/UMB-Heat-Cor.py
python fig8/RandomForestClassifier.py
python fig8/RandomForestRegressor.py

The figure scripts use AUR.csv or UMB.csv; the figure 6 and figure 8 scripts use combined.csv.

Citation

If you use this code or data, please cite the paper and the dataset.

Paper

Shakiba, M., Rassouli, F.B. Joining up the scattered anticancer knowledge on
auraptene and umbelliprenin: a meta-analysis. Sci Rep 14, 11770 (2024).
https://doi.org/10.1038/s41598-024-62747-z

Dataset

Shakiba, M., Rassouli, F.B. (2024). Data from: Joining up the scattered anticancer
knowledge on auraptene and umbelliprenin: A meta-analysis [Dataset]. Dryad.
https://doi.org/10.5061/dryad.cvdncjt9k