Provide AI-ready drug-discovery datasets and benchmarks from PyTDC for AI/ML modeling.
npx skills add https://github.com/jackspace/ClaudeSkillz --skill pytdc
Please help me install this Agent Skill. Skill: pytdc Source: https://github.com/jackspace/ClaudeSkillz/tree/main/skills/scientific-pkg-pytdc Command: npx skills add https://github.com/jackspace/ClaudeSkillz --skill pytdc
PyTDC provides AI-ready therapeutics datasets and benchmarks (ADME, toxicity, DTI) with standardized splits, benchmarks, and data processing utilities to accelerate drug-discovery research and evaluation.
High-intent search queries and answers about installing and using this skill.
PyTDC provides curated, AI-ready drug-discovery datasets including ADME, toxicity, DTI, and molecular generation tasks with standardized splits and evaluation metrics. Install via pip install PyTDC, then use data = ADME(name='Caco2_Wang'); split = data.get_split(method='scaffold'); df = data.get_data(format='df') to retrieve DataFrames for model training and evaluation.
PyTDC outputs standardized DataFrames and supports multiple partitioning strategies: scaffold, random, and cold splits. Data access functions (get_split, get_data) enable flexible retrieval across ADME, toxicity, DTI, DDI, HTS, QM, and molecular generation workflows with consistent benchmark protocols.
Yes, PyTDC supports molecule generation and optimization workflows through molecular generation oracles and benchmarking utilities. Beyond single-instance prediction, it handles multi-instance tasks and provides datasets for pharmacokinetic property prediction and molecular optimization evaluation.
PyTDC requires Python and the PyTDC package (pip install PyTDC). No additional dependencies are required. You need domain knowledge of drug-discovery metrics (ADME, toxicity, DTI) and familiarity with pandas DataFrames for data manipulation and model benchmarking workflows.
Yes, PyTDC includes standardized scaffold, random, and cold partition splits across all datasets to ensure reproducible benchmarking. These splits enable consistent model evaluation and comparison across drug-discovery prediction and generation tasks.
PyTDC datasets cover ADME (absorption, distribution, metabolism, excretion), toxicity prediction, DTI (drug-target interaction), DDI (drug-drug interaction), HTS (high-throughput screening), QM (quantum mechanics), and molecular generation benchmarks for single and multi-instance ML modeling.
robinbarvaagAccesses PyTDC AI-ready drug discovery datasets and standardized benchmarks for ADME, toxicity, DTI, and molecular generation tasks.
yf8578Access curated drug discovery datasets and benchmarks for model training and evaluation.
Yezez9Convert SMILES strings and RDKit molecules into ML-ready feature representations.
jaechang-hitsLoad PyTDC drug discovery datasets with standardized splits and metrics.
