drug-likeness

De novo molecular design with deep molecular generative models for protein-protein interaction(PPI) inhibitors

Drug-Likeness

This project implements a Graph Neural Network (GNN) model for classifying molecules in the context of Traditional Chinese Medicine drug-likeness. The model uses molecular graph representations and deep learning techniques to predict whether a given molecule is druglike

Explore drug-like space with deep generative models

RDKit-based workflow for interpretable small-molecule developability assessment, drug-likeness screening, and reproducible cheminformatics analysis.

An intelligent machine learning application for predicting the drug-likeness of molecular compounds using molecular descriptors and predictive models. Includes real-time inference, batch analysis, interactive visualizations, REST API integration, and exportable reports for research and screening workflows.

Code used in the elective course Advanced Computational Methods in Drug Discovery: AI and Physics-Based Simulations at Leiden University.

Drug likeness screening tool using RDKit molecular descriptors and Lipinski's Rule of Five.

Automated chemical compound profiling, Lipinski Rule of 5 drug-likeness assessment, GHS hazard classification, and interactive 3D molecular visualization using PubChem REST APIs.