AI Skill MartBrowse skillsPromptsProgramsLearnWhat is a Skill?
 
← Back to list

PyTDC (Therapeutics Data Commons) Workflow

A disciplined workflow for PyTDC 1.1.15: metadata-first discovery, approval-gated dataset downloads, task-aware splits, exact evaluator names, benchmark groups, and bounded molecular oracles.

Data & AnalyticsAdvanced★ 33,030⑂ 3,248AI score 8/10Last updated: Aug 9, 2026

What it does

  • Explores the Therapeutics Data Commons registry (tasks, datasets, benchmarks, evaluators, oracles) without downloading anything.
  • Gives a reproducible install: isolated uv + CPython 3.11 env with PyTDC==1.1.15 and setuptools==80.9.0 pinned (PyTDC still imports legacy pkg_resources).
  • Enforces a discover → plan → ask → execute policy; downloads only happen with --execute (plus --download for oracle checkpoints and MolGen corpora).
  • Documents real split behavior and caveats (random, scaffold, cold_split, time), including that the 1.1.15 random splitter uses a fixed random_state=1 for validation.
  • Pins down easily-missed API facts: exact metric names (PCC, ROC-AUC, micro-f1), specialized benchmark-group classes, and evaluate_many needing ≥5 prediction dicts.
  • Bundled CLIs for metadata discovery, load/split, benchmark evaluation, oracle scoring, and read-only cache auditing.

Who it's for

  • Therapeutic ML researchers and data scientists working on ADME, Tox, DTI, DDI, or molecule generation.
  • Teams that need reproducible TDC leaderboard submissions.
  • Environments where dataset licensing, disk usage, and network egress must be reviewed before any download.

Examples

  1. "List ADME datasets without downloading anything" → discover_metadata.py --kind datasets --task ADME
  2. "Plan a scaffold split of Caco2_Wang with seed 42, then run it after I approve" → load_and_split_data.py ... --execute
  3. "Score a few SMILES with the local QED oracle" → molecular_generation.py score --oracle QED --smiles CCO
  4. "Validate my admet_group prediction JSON before evaluating" → benchmark_evaluation.py

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/K-Dense-AI/scientific-agent-skills/HEAD/skills/pytdc/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on AI Skill Mart.
Copy the skills/pytdc folder from the GitHub repo K-Dense-AI/scientific-agent-skills into my ~/.claude/skills/pytdc/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/K-Dense-AI/scientific-agent-skills.git && mkdir -p ~/.claude/skills && cp -r scientific-agent-skills/skills/pytdc ~/.claude/skills/pytdc

⚠ This is a third-party skill. Check the source repository before installing.

  1. Clone the repository: git clone https://github.com/K-Dense-AI/scientific-agent-skills.git
  2. Copy the skill into Claude's skills folder: mkdir -p ~/.claude/skills && cp -r scientific-agent-skills/skills/pytdc ~/.claude/skills/pytdc
  3. Install uv if you don't have it: curl -LsSf https://astral.sh/uv/install.sh | sh
  4. Create the isolated environment: uv venv --python 3.11 .venv-pytdc
  5. Preview the footprint first: uv pip install --dry-run --python .venv-pytdc/bin/python "setuptools==80.9.0" "PyTDC==1.1.15" (~123 packages, hundreds of MB).
  6. Install for real by rerunning the same command without --dry-run.
  7. Restart Claude Code and try: "Use the PyTDC skill to list ADME datasets."
  8. When Claude asks permission to download a dataset or oracle checkpoint, check the dataset license, size, and cache path before approving.
View source on GitHub ↗License: MIT