Pytdc

Need clean, benchmark-ready datasets for drug discovery ML? This gives you instant access to TDC's curated collections covering ADME properties, toxicity, drug-target interactions, and molecular generation tasks. The standardized interface means you can load any dataset with the same three-line pattern, and it handles the annoying stuff like scaffold splits and cold-split protocols that pharma ML actually needs. Comes with 22 datasets in the ADMET benchmark group alone, plus molecular oracles for property optimization. If you're tired of scraping and cleaning pharmacological data from ten different sources with inconsistent formats, this is the obvious choice. The scaffold splitting and proper train-test separation are worth it alone.

skills.sh installs: 275
GitHub stars: 33.2K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add davila7/claude-code-templates --skill pytdc --agent claude-code

Installs into .claude/skills of the current project.

Files

PyTDC (Therapeutics Data Commons)

Overview

PyTDC is an open-science platform providing AI-ready datasets and benchmarks for drug discovery and development. Access curated datasets spanning the entire therapeutics pipeline with standardized evaluation metrics and meaningful data splits, organized into three categories: single-instance prediction (molecular/protein properties), multi-instance prediction (drug-target interactions, DDI), and generation (molecule generation, retrosynthesis).