A Euclidean diffusion model for structure-based drug design.

Kosmos: An AI Scientist for Autonomous Discovery - An implementation and adaptation to be driven by Claude Code or API - Based on the Kosmos AI Paper - https://arxiv.org/abs/2511.02824

Vision Bridge Transformer at Scale

Bridging Smoothed Molecular Dynamics and Score-Based Learning for Conformational Ensembles

Feynman-Kac steering for controllable protein diffusion

Official repository of "ODesign: A World Model for Biomolecular Interaction Design"

TorchCFM: a Conditional Flow Matching library

Code for running RFdiffusion

A general framework for inference-time scaling and steering of diffusion models with arbitrary rewards.

Flow Matching for Electron Redistribution

Code for "Calibrating Generative Models" by Henry Smith, Nathaniel Diamant, and Brian Trippe

Repository for StripedHyena, a state-of-the-art beyond Transformer architecture

Biological foundation modeling from molecular to genome scale

A lightweight, fast pocket finder in Python.

dLLM: Simple Diffusion Language Modeling

One protein is all you need

The Family of Diffusion Protein Language Models (DPLM)

The official implementation for Pseudo Numerical Methods for Diffusion Models on Manifolds (PNDM, PLMS | ICLR2022)

Discovering Interpretable Features in Protein Language Models via Sparse Autoencoders

[ICLR 2025] Code for the paper "Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning"

Lightning fast C++/CUDA neural network framework

Samples for CUDA Developers which demonstrates features in CUDA Toolkit

MMseqs2: ultra fast and sensitive search and clustering suite

Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

A fully open source biomolecular structure prediction model based on AlphaFold3

This is a curated list of research papers, resources and tools related to using Graph Neural Networks (GNNs) for drug discovery.

A tiny CPU simulator written in Python

Merkle tree implementation in Rust with configurable storage backends and hash functions. Fixed depth and incremental only. Optimized for fast proof generation.