chemprop

MoleculeNet benchmark dataset & MolMapNet dataset

Antibiotic discovery using graph deep learning, with Chemprop.

This repository contains code and documentation for participating in the OpenADMET + ExpansionRx Blind Challenge. The goal of this challenge is to develop machine learning models to predict various ADMET properties of small molecules using the provided dataset.

Hydrolysis reaction rate estimation from chemical structure

Multi-task learning (BBB + P-gp) for blood-brain barrier permeability prediction

This project is focused on the development and implementation of chemical molecular models following MolSSI's best practices for version control and Python package management. The project is intended for beginners who are interested in working with chemical molecular simulations.

Neural network for predicting Fukui indices using Kernel-based Attention Networks (KAN) with Chebyshev graph convolutions.

AI-driven ADMET property prediction platform utilizing Chemprop-RDKit models trained on Therapeutics Data Commons (TDC) datasets. Features fast molecular property evaluation, radar visual summaries, drug candidate profiling, and interactive web visualization tools.