DeepGraphLearning/torchdrug

Health v2 · maintenance only

Adoption not part of the score

1587 stars · 221 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

TorchDrug is a PyTorch-based machine learning library for drug discovery, covering graph neural networks, deep generative models, and reinforcement learning on molecular and biomedical data. It provides tensor-based graph operations, common datasets, benchmarks, and scalable multi-GPU training for rapid prototyping of drug discovery models.

Use cases

• predict molecular properties with graph neural networks
• generate novel molecules with deep generative models
• predict chemical reactions and retrosynthesis routes
• reason over biomedical knowledge graphs
• learn protein and molecule representations with pretrained models
• run graph machine learning research in PyTorch

When to choose

• you want to apply deep learning to drug discovery with minimal chemistry domain knowledge
• you need ready-made molecular datasets, GNN layers, and benchmarks in PyTorch
• you want scalable training across CPUs, GPUs, or distributed settings

When to avoid

• you need a production cheminformatics toolkit rather than a research library
• your project requires Python > 3.10 or very recent PyTorch versions
• you need non-graph machine learning unrelated to molecules or biomedical data

Facets

library · maturity active

machine-learning deep-learning graphics data-science benchmarking machine-learning deep-learning bioinformatics healthcare python windows pytorch graph-neural-networks drug-discovery molecules retrosynthesis knowledge-graphs molecule-generation protein-representation algorithms linux macos gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="DeepGraphLearning/torchdrug")