DyNaMo

DyNaMo is a fine‑tuned Random Forest machine learning framework developed to predict nanoparticle formation with FDA recognized dyes in drug-excipient nanoparticle systems. The repository is organized into two main components: DyNaMo modeling and PCA‑based exploratory analysis.

---

Overview

Nanoparticle formation in dye‑based systems depends on complex physiochemical relationships between drug and dye mooecules. DyNaMo combines:

• Random Forest modeling for prediction
• Principal Component Analysis (PCA) for data exploration and dimensionality reduction

to provide both predictive performance and scientific interpretability.

---

Repository Structure

DyNaMo-RF/
│
├── DyNaMo/
│   ├── DyNaMo_Data/                    # Input data for DyNaMo modeling
│   └── DyNaMo_Predictions_RF.ipynb     # Random Forest training and prediction notebook
│
├── PCA/
│   ├── PCA_Data/                       # Data used for PCA analysis
│   └── Excip_Dye_PCA.ipynb             # PCA of excipient–dye systems
│
└── README.md                           # Project documentation

---

Features

• Fine‑tuned Random Forest model for nanoparticle formation prediction
• Focus on FDA recognized dyes
• PCA for exploratory data analysis and pattern discovery in excipient chemical space
• Interpretable outputs supporting experimental insight
• Designed for nanomaterials and dye‑based systems research

Requirements and Installation:

This project requires Python 3.10.2 and relies on machine learning and cheminformatics libraries. RDKit is a core dependency and is best installed using conda.

---

System Requirements

Python 3.10.2 Jupyter Notebook or JupyterLab

Required Packages

rdkit == 2022.3.4

numpy

pandas

scikit‑learn

matplotlib

seaborn