# Disease-Pathway-Drug Proximity Analysis and Network Visualization

This repository provides a pipeline to **prioritize drug candidates** for repurposing based on their network-based proximity to biological pathways associated with a disease. 

Additionally, it includes a utility to visualize the tripartite relationship between **drugs**, **genes**, and **diseases**.

---

## 🗂️ Project Structure

```
.
├── code/
│   ├── drug_pathway_proximity/               # 💻 Module for running proximity analysis 
│   ├── plot_drug_gene_disease_network.py     # 🔍 Plotting script for tripartite graphs
├── data/                                     # 📂 Input CSV/TSV files (in module docs)
├── results/                                  # 📂 Proximity results in CSV
├── figures/                                  # 📂 Network optional output plots
```

---

## 🔎 Functionality Summary

### 🔬 1. Drug Pathway Proximity Module

Python module to execute proximity analysis between pathways/diseases and drugs in the interactome.

The core pipeline computes **minimum SPL**, **average SPL**, and **network-based proximity Z-score** between:

- `S`: Genes shared between a **disease** and a **pathway**
- `T`: Target genes of a **drug**

These metrics are based on the methodology in:

> *Guney, E. et al. "Network-based in silico drug efficacy screening". Nature Communications 7, 10331 (2016).*  
> https://www.nature.com/articles/ncomms10331

The results include:
- Closest interacting gene pairs
- Intermediate nodes in shortest paths
- Degree-preserving z-score evaluation of proximity

---

📂 For a full description of the proximity module, see [`drug_pathway_proximity/README.md`](code/drug_pathway_proximity/README.md).

---

### 🧬 2. Tripartite Network Visualization

The file `plot_drug_gene_disease_network.py` visualizes the relationships between:
- Drugs
- Genes (present in a related pathway)
- Diseases

The network is tripartite and uses distinct color coding for:
- Drugs
- Pathway-associated genes
- Diseases

The relationship between drugs and genes is colored based on the underlying regulation (up or down).

---

## ▶️ How to Run the Pipeline in the Drug Pathway Proximity Module

From the `code/` folder:

```bash
cd code
python -m drug_pathway_proximity
```

This will generate a CSV in `../results/drug_pathway_proximity_results.csv` containing all proximity metrics and relationships.

---

## 📈 How to Plot a Tripartite Network

After generating results, you can plot a **Drug–Gene–Disease** network for selected triplets using:

```bash
python plot_drug_gene_disease_network.py
```

This script takes the `tripartite_network.xlsx` file as input. 

It outputs `.png` or `.pdf` figures to `../figures/` showing the interaction context for each candidate.

---

## ⚙️ Environment Setup

Install requirements using conda:

```bash
conda env create -f environment.yml
conda activate drug-pathway-proximity
```

---

## 👩‍🔬 Authors and Credits

- B. Otero-Carrasco, L. Prieto-Santamaría, and A. Rodríguez-González \
  MEDAL · [https://medal.ctb.upm.es](https://medal.ctb.upm.es)  
  CTB · [https://ctb.upm.es](https://ctb.upm.es) 

- **Lucía Prieto Santamaría**  
  Assistant Professor, [Universidad Politécnica de Madrid (UPM)](https://www.upm.es/)  
  Main developer of this code.

This work builds upon previous DREBIOP studies:
- Otero-Carrasco et al. *BMC Genomics*, 2024 – [10.1186/s12864-023-09913-1](https://doi.org/10.1186/s12864-023-09913-1)
- Otero Carrasco et al., *RExPO 2024* – [10.58647/REXPO.24000039.v1](https://doi.org/10.58647/REXPO.24000039.v1)

---