Provide standardized drug discovery datasets with train/validation/test splits.
npx skills add https://github.com/ovachiever/droid-tings --skill pytdc-ovachiever
Please help me install this Agent Skill. Skill: pytdc Source: https://github.com/ovachiever/droid-tings/tree/main/skills/pytdc Command: npx skills add https://github.com/ovachiever/droid-tings --skill pytdc-ovachiever
PyTDC provides AI-ready datasets and benchmarks for drug discovery, offering standardized train/validation/test splits and meaningful evaluation metrics across ADME, Toxicity, Multi-instance, and Generation tasks.
Example: load an ADME task and obtain scaffold splits, then access train/valid/test partitions.
High-intent search queries and answers about installing and using this skill.
PyTDC delivers AI-ready datasets for drug discovery with consistent train/validation/test splits across ADME, toxicity, DTI, and generation tasks. Access pre-processed molecular data with standardized benchmarks and evaluation metrics for reproducible model assessment.
PyTDC covers single-instance prediction (ADME, toxicity, HTS, QM), multi-instance interaction tasks (DTI, DDI, PPI), and generation workflows (MolGen, RetroSyn, PairMolGen). Each task includes scaffold, cold, and temporal split strategies for rigorous evaluation.
Yes. PyTDC provides standardized benchmarks with consistent evaluation interfaces and metrics across datasets. Use built-in train/validation/test partitions and metadata to assess model performance reproducibly on ADMET and other property prediction tasks.
PyTDC offers scaffold, cold, and temporal splits alongside standard random splits. Load any dataset, specify your split strategy, and retrieve train/validation/test partitions with corresponding metadata to support rigorous model validation.
PyTDC requires numpy and pandas alongside the tdc package itself. These dependencies enable efficient data manipulation, array operations, and integration with standard machine learning and deep learning frameworks for drug discovery modeling.
PyTDC's standardized splits (scaffold, cold, temporal) control for data leakage specific to drug discovery—preventing similar molecules or temporal information from spanning train/test boundaries. This ensures realistic model evaluation compared to random partitioning alone.
InternScienceConvert IUPAC chemical names to SMILES strings via OPSIN.
Estrella-231Load PyTDC therapeutic datasets with scaffold and cold splits for ML modeling.
dkyazzentwatwaCompare machine learning models using cross-validation and performance metrics.
jackspaceProvide AI-ready drug-discovery datasets and benchmarks from PyTDC for AI/ML modeling.
