Is PyTDC compatible with graph neural networks?

The skill provides utilities for converting chemical SMILES into graph representations compatible with PyTorch Geometric (PyG) and DGL.

What is PyTDC in the context of Claude Code?

PyTDC (Therapeutics Data Commons) is a skill that enables Claude to assist with drug discovery tasks by providing standardized access to datasets, benchmarks, and molecular oracles.

Does PyTDC support molecular generation tasks?

Yes, the skill includes guidance on using molecular oracles to optimize specific chemical properties during de novo molecule generation.

Can this skill help with molecular property prediction?

Yes, it provides curated datasets and workflows for predicting ADME (Absorption, Distribution, Metabolism, Excretion) and toxicity properties.

How does this skill handle chemical data splits?

It implements specialized splitting strategies like scaffold splits for chemical diversity and cold-target splits for drug-target interaction tasks.

PyTDC Therapeutics AI

Name: PyTDC Therapeutics AI
Author: K-Dense-AI

byK-Dense-AI

•

3,719

データサイエンスとML

Facilitates drug discovery and therapeutic machine learning with curated datasets, ADMET benchmarks, and molecular optimization oracles.

概要

The PyTDC skill equips Claude with specialized knowledge for the Therapeutics Data Commons, a comprehensive open-science platform for pharmaceutical machine learning. It provides standardized access to diverse datasets spanning the entire therapeutics pipeline, from molecular property prediction (ADME-Tox) to drug-target interactions and de novo molecular generation. This skill is ideal for researchers and developers needing robust benchmarking, specialized chemical data splits like scaffold or cold-start, and property-optimizing oracles to accelerate the development of AI-driven therapeutic solutions.

主な機能

Standardized access to 70+ curated therapeutics and drug discovery datasets
Specialized data splitting strategies including scaffold, cold-drug, and temporal splits
3,719 GitHub stars
Support for multi-entity tasks including drug-target and drug-drug interactions
Integration with 17+ molecular oracles for goal-directed property optimization
Comprehensive ADMET benchmark groups for systematic model evaluation

ユースケース

Building and evaluating ML models for pharmacokinetic (ADMET) property prediction
Conducting goal-directed molecular generation for lead compound optimization
Benchmarking pharmacological prediction models against industry-standard therapeutic datasets

概要

主な機能

Standardized access to 70+ curated therapeutics and drug discovery datasets
Specialized data splitting strategies including scaffold, cold-drug, and temporal splits
3,719 GitHub stars
Support for multi-entity tasks including drug-target and drug-drug interactions
Integration with 17+ molecular oracles for goal-directed property optimization
Comprehensive ADMET benchmark groups for systematic model evaluation

ユースケース

Building and evaluating ML models for pharmacokinetic (ADMET) property prediction
Conducting goal-directed molecular generation for lead compound optimization
Benchmarking pharmacological prediction models against industry-standard therapeutic datasets