Project

GNN-Based Pharmacological Interaction Engine

Graph neural model predicting drug interaction severity using SMILES-based molecular graphs and RDKit descriptors. Benchmarked GAT, GIN, and MPNN variants across 16,000+ structures.

GitHub

graph-neural-networksdrug-discoverybioinformatics

Problem

Predicting drug-drug interactions requires understanding complex molecular structure relationships that traditional ML methods struggle with.

Solution

Graph neural networks (GAT, GIN, MPNN) operating on SMILES-based molecular graphs augmented with RDKit chemical descriptors for severity prediction.

Impact

Open-source benchmark comparing GNN architectures on pharmacological interaction prediction at scale.

Stack

PyTorchPyTorch GeometricRDKitPythonScikit-learn